5 Changes between 0.9.6 and 0.9.6a [xx XXX 2001]
7 *) Fix PKCS#7 decode routines so they correctly update the length
8 after reading an EOC for the EXPLICIT tag.
10 [This change does not apply to 0.9.7.]
12 *) Fix bug in PKCS#12 key generation routines. This was triggered
13 if a 3DES key was generated with a 0 initial byte. Include
14 PKCS12_BROKEN_KEYGEN compilation option to retain the old
15 (but broken) behaviour.
18 *) Fix memory leaks in err.c: free err_data string if necessary;
19 don't write to the wrong index in ERR_set_error_data.
22 *) Implement ssl23_peek (analogous to ssl23_read), which previously
26 *) Replace rdtsc with _emit statements for VC++ version 5.
27 [Jeremy Cooper <jeremy@baymoo.org>]
29 *) Make it possible to reuse SSLv2 sessions.
32 *) In copy_email() check for >= 0 as a return value for
33 X509_NAME_get_index_by_NID() since 0 is a valid index.
34 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>]
36 *) Avoid coredump with unsupported or invalid public keys by checking if
37 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
38 PKCS7_verify() fails with non detached data.
41 *) Don't use getenv in library functions when run as setuid/setgid.
42 New function OPENSSL_issetugid().
45 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
46 due to incorrect handling of multi-threading:
48 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
50 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
52 3. Count how many times MemCheck_off() has been called so that
53 nested use can be treated correctly. This also avoids
54 inband-signalling in the previous code (which relied on the
55 assumption that thread ID 0 is impossible).
58 *) Add "-rand" option also to s_client and s_server.
61 *) Fix CPU detection on Irix 6.x.
62 [Kurt Hockenbury <khockenb@stevens-tech.edu> and
63 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
65 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME
68 [This change does not apply to 0.9.7.]
70 *) Use the cached encoding of an X509_NAME structure rather than
71 copying it. This is apparently the reason for the libsafe "errors"
72 but the code is actually correct.
75 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
76 Bleichenbacher's DSA attack.
77 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
78 to be set and top=0 forces the highest bit to be set; top=-1 is new
79 and leaves the highest bit random.
80 [Ulf Moeller, Bodo Moeller]
82 *) In the NCONF_...-based implementations for CONF_... queries
83 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
84 a temporary CONF structure with the data component set to NULL
85 (which gives segmentation faults in lh_retrieve).
86 Instead, use NULL for the CONF pointer in CONF_get_string and
87 CONF_get_number (which may use environment variables) and directly
88 return NULL from CONF_get_section.
91 *) Fix potential buffer overrun for EBCDIC.
94 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign
95 keyUsage if basicConstraints absent for a CA.
98 *) Make SMIME_write_PKCS7() write mail header values with a format that
99 is more generally accepted (no spaces before the semicolon), since
100 some programs can't parse those values properly otherwise. Also make
101 sure BIO's that break lines after each write do not create invalid
105 *) Make the CRL encoding routines work with empty SEQUENCE OF. The
106 macros previously used would not encode an empty SEQUENCE OF
107 and break the signature.
109 [This change does not apply to 0.9.7.]
111 *) Zero the premaster secret after deriving the master secret in
115 *) Add some EVP_add_digest_alias registrations (as found in
116 OpenSSL_add_all_digests()) to SSL_library_init()
117 aka OpenSSL_add_ssl_algorithms(). This provides improved
118 compatibility with peers using X.509 certificates
119 with unconventional AlgorithmIdentifier OIDs.
122 *) Fix for Irix with NO_ASM.
123 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
125 *) ./config script fixes.
126 [Ulf Moeller, Richard Levitte]
128 *) Fix 'openssl passwd -1'.
131 *) Change PKCS12_key_gen_asc() so it can cope with non null
132 terminated strings whose length is passed in the passlen
133 parameter, for example from PEM callbacks. This was done
134 by adding an extra length parameter to asc2uni().
135 [Steve Henson, reported by <oddissey@samsung.co.kr>]
137 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
138 call failed, free the DSA structure.
141 *) Fix to uni2asc() to cope with zero length Unicode strings.
142 These are present in some PKCS#12 files.
145 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
146 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
147 when writing a 32767 byte record.
148 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>]
150 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c),
151 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}.
153 (RSA objects have a reference count access to which is protected
154 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
155 so they are meant to be shared between threads.)
156 [Bodo Moeller, Geoff Thorpe; original patch submitted by
157 "Reddie, Steven" <Steven.Reddie@ca.com>]
159 *) Fix a deadlock in CRYPTO_mem_leaks().
162 *) Use better test patterns in bntest.
165 *) rand_win.c fix for Borland C.
168 *) BN_rshift bugfix for n == 0.
171 *) Add a 'bctest' script that checks for some known 'bc' bugs
172 so that 'make test' does not abort just because 'bc' is broken.
175 *) Store verify_result within SSL_SESSION also for client side to
176 avoid potential security hole. (Re-used sessions on the client side
177 always resulted in verify_result==X509_V_OK, not using the original
178 result of the server certificate verification.)
181 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type
182 SSL3_RT_APPLICATION_DATA, return 0.
183 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
187 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
188 releases, have been re-implemented by renaming the previous
189 implementations of ssl2_read and ssl3_read to ssl2_read_internal
190 and ssl3_read_internal, respectively, and adding 'peek' parameters
191 to them. The new ssl[23]_{read,peek} functions are calls to
192 ssl[23]_read_internal with the 'peek' flag set appropriately.
193 A 'peek' parameter has also been added to ssl3_read_bytes, which
194 does the actual work for ssl3_read_internal.
197 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
198 the method-specific "init()" handler. Also clean up ex_data after
199 calling the method-specific "finish()" handler. Previously, this was
200 happening the other way round.
203 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
204 The previous value, 12, was not always sufficient for BN_mod_exp().
207 *) Make sure that shared libraries get the internal name engine with
208 the full version number and not just 0. This should mark the
209 shared libraries as not backward compatible. Of course, this should
210 be changed again when we can guarantee backward binary compatibility.
213 *) Fix typo in get_cert_by_subject() in by_dir.c
214 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>]
216 *) Rework the system to generate shared libraries:
218 - Make note of the expected extension for the shared libraries and
219 if there is a need for symbolic links from for example libcrypto.so.0
220 to libcrypto.so.0.9.7. There is extended info in Configure for
223 - Make as few rebuilds of the shared libraries as possible.
225 - Still avoid linking the OpenSSL programs with the shared libraries.
227 - When installing, install the shared libraries separately from the
231 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument.
233 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
234 and not in SSL_clear because the latter is also used by the
235 accept/connect functions; previously, the settings made by
236 SSL_set_read_ahead would be lost during the handshake.
237 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>]
239 *) Correct util/mkdef.pl to be selective about disabled algorithms.
240 Previously, it would create entries for disableed algorithms no
244 *) Added several new manual pages for SSL_* function.
247 Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
249 *) In ssl23_get_client_hello, generate an error message when faced
250 with an initial SSL 3.0/TLS record that is too small to contain the
251 first two bytes of the ClientHello message, i.e. client_version.
252 (Note that this is a pathologic case that probably has never happened
253 in real life.) The previous approach was to use the version number
254 from the record header as a substitute; but our protocol choice
255 should not depend on that one because it is not authenticated
256 by the Finished messages.
259 *) More robust randomness gathering functions for Windows.
260 [Jeffrey Altman <jaltman@columbia.edu>]
262 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
263 not set then we don't setup the error code for issuer check errors
264 to avoid possibly overwriting other errors which the callback does
265 handle. If an application does set the flag then we assume it knows
266 what it is doing and can handle the new informational codes
270 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
271 a general "ANY" type, as such it should be able to decode anything
272 including tagged types. However it didn't check the class so it would
273 wrongly interpret tagged types in the same way as their universal
274 counterpart and unknown types were just rejected. Changed so that the
275 tagged and unknown types are handled in the same way as a SEQUENCE:
276 that is the encoding is stored intact. There is also a new type
277 "V_ASN1_OTHER" which is used when the class is not universal, in this
278 case we have no idea what the actual type is so we just lump them all
282 *) On VMS, stdout may very well lead to a file that is written to
283 in a record-oriented fashion. That means that every write() will
284 write a separate record, which will be read separately by the
285 programs trying to read from it. This can be very confusing.
287 The solution is to put a BIO filter in the way that will buffer
288 text until a linefeed is reached, and then write everything a
289 line at a time, so every record written will be an actual line,
290 not chunks of lines and not (usually doesn't happen, but I've
291 seen it once) several lines in one record. BIO_f_linebuffer() is
294 Currently, it's a VMS-only method, because that's where it has
295 been tested well enough.
298 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
299 it can return incorrect results.
300 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
301 but it was in 0.9.6-beta[12].)
304 *) Disable the check for content being present when verifying detached
305 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
306 include zero length content when signing messages.
309 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
310 BIO_ctrl (for BIO pairs).
313 *) Add DSO method for VMS.
316 *) Bug fix: Montgomery multiplication could produce results with the
320 *) Add RPM specification openssl.spec and modify it to build three
321 packages. The default package contains applications, application
322 documentation and run-time libraries. The devel package contains
323 include files, static libraries and function documentation. The
324 doc package contains the contents of the doc directory. The original
325 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
328 *) Add a large number of documentation files for many SSL routines.
329 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
331 *) Add a configuration entry for Sony News 4.
332 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>]
334 *) Don't set the two most significant bits to one when generating a
335 random number < q in the DSA library.
338 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
339 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
340 the underlying transport is blocking) if a handshake took place.
341 (The default behaviour is needed by applications such as s_client
342 and s_server that use select() to determine when to use SSL_read;
343 but for applications that know in advance when to expect data, it
344 just makes things more complicated.)
347 *) Add RAND_egd_bytes(), which gives control over the number of bytes read
351 *) Add a few more EBCDIC conditionals that make `req' and `x509'
352 work better on such systems.
353 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
355 *) Add two demo programs for PKCS12_parse() and PKCS12_create().
356 Update PKCS12_parse() so it copies the friendlyName and the
357 keyid to the certificates aux info.
360 *) Fix bug in PKCS7_verify() which caused an infinite loop
361 if there was more than one signature.
362 [Sven Uszpelkat <su@celocom.de>]
364 *) Major change in util/mkdef.pl to include extra information
365 about each symbol, as well as presentig variables as well
366 as functions. This change means that there's n more need
367 to rebuild the .num files when some algorithms are excluded.
370 *) Allow the verify time to be set by an application,
371 rather than always using the current time.
374 *) Phase 2 verify code reorganisation. The certificate
375 verify code now looks up an issuer certificate by a
376 number of criteria: subject name, authority key id
377 and key usage. It also verifies self signed certificates
378 by the same criteria. The main comparison function is
379 X509_check_issued() which performs these checks.
381 Lot of changes were necessary in order to support this
382 without completely rewriting the lookup code.
384 Authority and subject key identifier are now cached.
386 The LHASH 'certs' is X509_STORE has now been replaced
387 by a STACK_OF(X509_OBJECT). This is mainly because an
388 LHASH can't store or retrieve multiple objects with
391 As a result various functions (which were all internal
392 use only) have changed to handle the new X509_STORE
393 structure. This will break anything that messed round
394 with X509_STORE internally.
396 The functions X509_STORE_add_cert() now checks for an
397 exact match, rather than just subject name.
399 The X509_STORE API doesn't directly support the retrieval
400 of multiple certificates matching a given criteria, however
401 this can be worked round by performing a lookup first
402 (which will fill the cache with candidate certificates)
403 and then examining the cache for matches. This is probably
404 the best we can do without throwing out X509_LOOKUP
405 entirely (maybe later...).
407 The X509_VERIFY_CTX structure has been enhanced considerably.
409 All certificate lookup operations now go via a get_issuer()
410 callback. Although this currently uses an X509_STORE it
411 can be replaced by custom lookups. This is a simple way
412 to bypass the X509_STORE hackery necessary to make this
413 work and makes it possible to use more efficient techniques
414 in future. A very simple version which uses a simple
415 STACK for its trusted certificate store is also provided
416 using X509_STORE_CTX_trusted_stack().
418 The verify_cb() and verify() callbacks now have equivalents
419 in the X509_STORE_CTX structure.
421 X509_STORE_CTX also has a 'flags' field which can be used
422 to customise the verify behaviour.
425 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
426 excludes S/MIME capabilities.
429 *) When a certificate request is read in keep a copy of the
430 original encoding of the signed data and use it when outputing
431 again. Signatures then use the original encoding rather than
432 a decoded, encoded version which may cause problems if the
433 request is improperly encoded.
436 *) For consistency with other BIO_puts implementations, call
437 buffer_write(b, ...) directly in buffer_puts instead of calling
440 In BIO_puts, increment b->num_write as in BIO_write.
441 [Peter.Sylvester@EdelWeb.fr]
443 *) Fix BN_mul_word for the case where the word is 0. (We have to use
444 BN_zero, we may not return a BIGNUM with an array consisting of
448 *) Avoid calling abort() from within the library when problems are
449 detected, except if preprocessor symbols have been defined
450 (such as REF_CHECK, BN_DEBUG etc.).
453 *) New openssl application 'rsautl'. This utility can be
454 used for low level RSA operations. DER public key
455 BIO/fp routines also added.
458 *) New Configure entry and patches for compiling on QNX 4.
459 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
461 *) A demo state-machine implementation was sponsored by
462 Nuron (http://www.nuron.com/) and is now available in
466 *) New options added to the 'dgst' utility for signature
467 generation and verification.
470 *) Unrecognized PKCS#7 content types are now handled via a
471 catch all ASN1_TYPE structure. This allows unsupported
472 types to be stored as a "blob" and an application can
473 encode and decode it manually.
476 *) Fix various signed/unsigned issues to make a_strex.c
478 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
480 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
481 length if passed a buffer. ASN1_INTEGER_to_BN failed
482 if passed a NULL BN and its argument was negative.
483 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
485 *) Modification to PKCS#7 encoding routines to output definite
486 length encoding. Since currently the whole structures are in
487 memory there's not real point in using indefinite length
488 constructed encoding. However if OpenSSL is compiled with
489 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
492 *) Added BIO_vprintf() and BIO_vsnprintf().
495 *) Added more prefixes to parse for in the the strings written
496 through a logging bio, to cover all the levels that are available
497 through syslog. The prefixes are now:
499 PANIC, EMERG, EMR => LOG_EMERG
500 ALERT, ALR => LOG_ALERT
501 CRIT, CRI => LOG_CRIT
502 ERROR, ERR => LOG_ERR
503 WARNING, WARN, WAR => LOG_WARNING
504 NOTICE, NOTE, NOT => LOG_NOTICE
505 INFO, INF => LOG_INFO
506 DEBUG, DBG => LOG_DEBUG
508 and as before, if none of those prefixes are present at the
509 beginning of the string, LOG_ERR is chosen.
511 On Win32, the LOG_* levels are mapped according to this:
513 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
514 LOG_WARNING => EVENTLOG_WARNING_TYPE
515 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
519 *) Made it possible to reconfigure with just the configuration
520 argument "reconf" or "reconfigure". The command line arguments
521 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
522 and are retrieved from there when reconfiguring.
526 [Assar Westerlund <assar@sics.se>, Richard Levitte]
528 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
531 *) The obj_dat.pl script was messing up the sorting of object
532 names. The reason was that it compared the quoted version
533 of strings as a result "OCSP" > "OCSP Signing" because
534 " > SPACE. Changed script to store unquoted versions of
535 names and add quotes on output. It was also omitting some
536 names from the lookup table if they were given a default
537 value (that is if SN is missing it is given the same
538 value as LN and vice versa), these are now added on the
539 grounds that if an object has a name we should be able to
540 look it up. Finally added warning output when duplicate
541 short or long names are found.
544 *) Changes needed for Tandem NSK.
545 [Scott Uroff <scott@xypro.com>]
547 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
548 RSA_padding_check_SSLv23(), special padding was never detected
549 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
550 version rollback attacks was not effective.
552 In s23_clnt.c, don't use special rollback-attack detection padding
553 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
554 client; similarly, in s23_srvr.c, don't do the rollback check if
555 SSL 2.0 is the only protocol enabled in the server.
558 *) Make it possible to get hexdumps of unprintable data with 'openssl
559 asn1parse'. By implication, the functions ASN1_parse_dump() and
560 BIO_dump_indent() are added.
563 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
564 these print out strings and name structures based on various
565 flags including RFC2253 support and proper handling of
566 multibyte characters. Added options to the 'x509' utility
567 to allow the various flags to be set.
570 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
571 Also change the functions X509_cmp_current_time() and
572 X509_gmtime_adj() work with an ASN1_TIME structure,
573 this will enable certificates using GeneralizedTime in validity
577 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
578 negative public key encodings) on by default,
579 NO_NEG_PUBKEY_BUG can be set to disable it.
582 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
583 content octets. An i2c_ASN1_OBJECT is unnecessary because
584 the encoding can be trivially obtained from the structure.
587 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
588 not read locks (CRYPTO_r_[un]lock).
591 *) A first attempt at creating official support for shared
592 libraries through configuration. I've kept it so the
593 default is static libraries only, and the OpenSSL programs
594 are always statically linked for now, but there are
595 preparations for dynamic linking in place.
596 This has been tested on Linux and True64.
599 *) Randomness polling function for Win9x, as described in:
600 Peter Gutmann, Software Generation of Practically Strong
604 *) Fix so PRNG is seeded in req if using an already existing
608 *) New options to smime application. -inform and -outform
609 allow alternative formats for the S/MIME message including
610 PEM and DER. The -content option allows the content to be
611 specified separately. This should allow things like Netscape
612 form signing output easier to verify.
615 *) Fix the ASN1 encoding of tags using the 'long form'.
618 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
619 STRING types. These convert content octets to and from the
620 underlying type. The actual tag and length octets are
621 already assumed to have been read in and checked. These
622 are needed because all other string types have virtually
623 identical handling apart from the tag. By having versions
624 of the ASN1 functions that just operate on content octets
625 IMPLICIT tagging can be handled properly. It also allows
626 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
627 and ASN1_INTEGER are identical apart from the tag.
630 *) Change the handling of OID objects as follows:
632 - New object identifiers are inserted in objects.txt, following
633 the syntax given in objects.README.
634 - objects.pl is used to process obj_mac.num and create a new
636 - obj_dat.pl is used to create a new obj_dat.h, using the data in
639 This is currently kind of a hack, and the perl code in objects.pl
640 isn't very elegant, but it works as I intended. The simplest way
641 to check that it worked correctly is to look in obj_dat.h and
642 check the array nid_objs and make sure the objects haven't moved
643 around (this is important!). Additions are OK, as well as
644 consistent name changes.
647 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
650 *) Addition of the command line parameter '-rand file' to 'openssl req'.
651 The given file adds to whatever has already been seeded into the
652 random pool through the RANDFILE configuration file option or
653 environment variable, or the default random state file.
656 *) mkstack.pl now sorts each macro group into lexical order.
657 Previously the output order depended on the order the files
658 appeared in the directory, resulting in needless rewriting
662 *) Patches to make OpenSSL compile under Win32 again. Mostly
663 work arounds for the VC++ problem that it treats func() as
664 func(void). Also stripped out the parts of mkdef.pl that
665 added extra typesafe functions: these no longer exist.
668 *) Reorganisation of the stack code. The macros are now all
669 collected in safestack.h . Each macro is defined in terms of
670 a "stack macro" of the form SKM_<name>(type, a, b). The
671 DEBUG_SAFESTACK is now handled in terms of function casts,
672 this has the advantage of retaining type safety without the
673 use of additional functions. If DEBUG_SAFESTACK is not defined
674 then the non typesafe macros are used instead. Also modified the
675 mkstack.pl script to handle the new form. Needs testing to see
676 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
677 the default if no major problems. Similar behaviour for ASN1_SET_OF
681 *) When some versions of IIS use the 'NET' form of private key the
682 key derivation algorithm is different. Normally MD5(password) is
683 used as a 128 bit RC4 key. In the modified case
684 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
685 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
686 as the old Netscape_RSA functions except they have an additional
687 'sgckey' parameter which uses the modified algorithm. Also added
688 an -sgckey command line option to the rsa utility. Thanks to
689 Adrian Peck <bertie@ncipher.com> for posting details of the modified
690 algorithm to openssl-dev.
693 *) The evp_local.h macros were using 'c.##kname' which resulted in
694 invalid expansion on some systems (SCO 5.0.5 for example).
695 Corrected to 'c.kname'.
696 [Phillip Porch <root@theporch.com>]
698 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
699 a STACK of email addresses from a certificate or request, these look
700 in the subject name and the subject alternative name extensions and
701 omit any duplicate addresses.
704 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
705 This makes DSA verification about 2 % faster.
708 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
709 (meaning that now 2^5 values will be precomputed, which is only 4 KB
710 plus overhead for 1024 bit moduli).
711 This makes exponentiations about 0.5 % faster for 1024 bit
712 exponents (as measured by "openssl speed rsa2048").
715 *) Rename memory handling macros to avoid conflicts with other
717 Malloc => OPENSSL_malloc
718 Malloc_locked => OPENSSL_malloc_locked
719 Realloc => OPENSSL_realloc
723 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
724 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
728 [John Jarvie <jjarvie@newsguy.com>]
730 *) The type-safe stack code has been rejigged. It is now only compiled
731 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
732 by default all type-specific stack functions are "#define"d back to
733 standard stack functions. This results in more streamlined output
734 but retains the type-safety checking possibilities of the original
738 *) The STACK code has been cleaned up, and certain type declarations
739 that didn't make a lot of sense have been brought in line. This has
740 also involved a cleanup of sorts in safestack.h to more correctly
741 map type-safe stack functions onto their plain stack counterparts.
742 This work has also resulted in a variety of "const"ifications of
743 lots of the code, especially "_cmp" operations which should normally
744 be prototyped with "const" parameters anyway.
747 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
748 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
749 (The PRNG state consists of two parts, the large pool 'state' and 'md',
750 where all of 'md' is used each time the PRNG is used, but 'state'
751 is used only indexed by a cyclic counter. As entropy may not be
752 well distributed from the beginning, 'md' is important as a
753 chaining variable. However, the output function chains only half
754 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
755 all of 'md', and seeding with STATE_SIZE dummy bytes will result
756 in all of 'state' being rewritten, with the new values depending
757 on virtually all of 'md'. This overcomes the 80 bit limitation.)
760 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
761 the handshake is continued after ssl_verify_cert_chain();
762 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
763 can lead to 'unexplainable' connection aborts later.
764 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
766 *) Major EVP API cipher revision.
767 Add hooks for extra EVP features. This allows various cipher
768 parameters to be set in the EVP interface. Support added for variable
769 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
770 setting of RC2 and RC5 parameters.
772 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
775 Remove lots of duplicated code from the EVP library. For example *every*
776 cipher init() function handles the 'iv' in the same way according to the
777 cipher mode. They also all do nothing if the 'key' parameter is NULL and
778 for CFB and OFB modes they zero ctx->num.
780 New functionality allows removal of S/MIME code RC2 hack.
782 Most of the routines have the same form and so can be declared in terms
785 By shifting this to the top level EVP_CipherInit() it can be removed from
786 all individual ciphers. If the cipher wants to handle IVs or keys
787 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
790 Change lots of functions like EVP_EncryptUpdate() to now return a
791 value: although software versions of the algorithms cannot fail
792 any installed hardware versions can.
795 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
796 this option is set, tolerate broken clients that send the negotiated
797 protocol version number instead of the requested protocol version
801 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
802 i.e. non-zero for export ciphersuites, zero otherwise.
803 Previous versions had this flag inverted, inconsistent with
804 rsa_tmp_cb (..._TMP_RSA_CB).
805 [Bodo Moeller; problem reported by Amit Chopra]
807 *) Add missing DSA library text string. Work around for some IIS
808 key files with invalid SEQUENCE encoding.
811 *) Add a document (doc/standards.txt) that list all kinds of standards
812 and so on that are implemented in OpenSSL.
815 *) Enhance c_rehash script. Old version would mishandle certificates
816 with the same subject name hash and wouldn't handle CRLs at all.
817 Added -fingerprint option to crl utility, to support new c_rehash
821 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
824 *) Fix for SSL server purpose checking. Server checking was
825 rejecting certificates which had extended key usage present
826 but no ssl client purpose.
827 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
829 *) Make PKCS#12 code work with no password. The PKCS#12 spec
830 is a little unclear about how a blank password is handled.
831 Since the password in encoded as a BMPString with terminating
832 double NULL a zero length password would end up as just the
833 double NULL. However no password at all is different and is
834 handled differently in the PKCS#12 key generation code. NS
835 treats a blank password as zero length. MSIE treats it as no
836 password on export: but it will try both on import. We now do
837 the same: PKCS12_parse() tries zero length and no password if
838 the password is set to "" or NULL (NULL is now a valid password:
839 it wasn't before) as does the pkcs12 application.
842 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
843 perror when PEM_read_bio_X509_REQ fails, the error message must
844 be obtained from the error queue.
847 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
848 it in ERR_remove_state if appropriate, and change ERR_get_state
849 accordingly to avoid race conditions (this is necessary because
850 thread_hash is no longer constant once set).
853 *) Bugfix for linux-elf makefile.one.
856 *) RSA_get_default_method() will now cause a default
857 RSA_METHOD to be chosen if one doesn't exist already.
858 Previously this was only set during a call to RSA_new()
859 or RSA_new_method(NULL) meaning it was possible for
860 RSA_get_default_method() to return NULL.
863 *) Added native name translation to the existing DSO code
864 that will convert (if the flag to do so is set) filenames
865 that are sufficiently small and have no path information
866 into a canonical native form. Eg. "blah" converted to
867 "libblah.so" or "blah.dll" etc.
870 *) New function ERR_error_string_n(e, buf, len) which is like
871 ERR_error_string(e, buf), but writes at most 'len' bytes
872 including the 0 terminator. For ERR_error_string_n, 'buf'
874 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
876 *) CONF library reworked to become more general. A new CONF
877 configuration file reader "class" is implemented as well as a
878 new functions (NCONF_*, for "New CONF") to handle it. The now
879 old CONF_* functions are still there, but are reimplemented to
880 work in terms of the new functions. Also, a set of functions
881 to handle the internal storage of the configuration data is
882 provided to make it easier to write new configuration file
883 reader "classes" (I can definitely see something reading a
884 configuration file in XML format, for example), called _CONF_*,
885 or "the configuration storage API"...
887 The new configuration file reading functions are:
889 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
890 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
892 NCONF_default, NCONF_WIN32
894 NCONF_dump_fp, NCONF_dump_bio
896 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
897 NCONF_new creates a new CONF object. This works in the same way
898 as other interfaces in OpenSSL, like the BIO interface.
899 NCONF_dump_* dump the internal storage of the configuration file,
900 which is useful for debugging. All other functions take the same
901 arguments as the old CONF_* functions wth the exception of the
902 first that must be a `CONF *' instead of a `LHASH *'.
904 To make it easer to use the new classes with the old CONF_* functions,
905 the function CONF_set_default_method is provided.
908 *) Add '-tls1' option to 'openssl ciphers', which was already
909 mentioned in the documentation but had not been implemented.
910 (This option is not yet really useful because even the additional
911 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
914 *) Initial DSO code added into libcrypto for letting OpenSSL (and
915 OpenSSL-based applications) load shared libraries and bind to
916 them in a portable way.
917 [Geoff Thorpe, with contributions from Richard Levitte]
919 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
921 *) Make sure _lrotl and _lrotr are only used with MSVC.
923 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
924 (the default implementation of RAND_status).
926 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
927 to '-clrext' (= clear extensions), as intended and documented.
928 [Bodo Moeller; inconsistency pointed out by Michael Attili
931 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
932 was larger than the MD block size.
933 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
935 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
936 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
937 using the passed key: if the passed key was a private key the result
938 of X509_print(), for example, would be to print out all the private key
942 *) des_quad_cksum() byte order bug fix.
943 [Ulf Möller, using the problem description in krb4-0.9.7, where
944 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
946 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
948 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
950 *) For easily testing in shell scripts whether some command
951 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
952 returns with exit code 0 iff no command of the given name is available.
953 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
954 the output goes to stdout and nothing is printed to stderr.
955 Additional arguments are always ignored.
957 Since for each cipher there is a command of the same name,
958 the 'no-cipher' compilation switches can be tested this way.
960 ('openssl no-XXX' is not able to detect pseudo-commands such
961 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
964 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
967 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
968 is set; it will be thrown away anyway because each handshake creates
970 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
971 to parameters -- in previous versions (since OpenSSL 0.9.3) the
972 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
973 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
976 *) New s_client option -ign_eof: EOF at stdin is ignored, and
977 'Q' and 'R' lose their special meanings (quit/renegotiate).
978 This is part of what -quiet does; unlike -quiet, -ign_eof
979 does not suppress any output.
982 *) Add compatibility options to the purpose and trust code. The
983 purpose X509_PURPOSE_ANY is "any purpose" which automatically
984 accepts a certificate or CA, this was the previous behaviour,
985 with all the associated security issues.
987 X509_TRUST_COMPAT is the old trust behaviour: only and
988 automatically trust self signed roots in certificate store. A
989 new trust setting X509_TRUST_DEFAULT is used to specify that
990 a purpose has no associated trust setting and it should instead
991 use the value in the default purpose.
994 *) Fix the PKCS#8 DSA private key code so it decodes keys again
995 and fix a memory leak.
998 *) In util/mkerr.pl (which implements 'make errors'), preserve
999 reason strings from the previous version of the .c file, as
1000 the default to have only downcase letters (and digits) in
1001 automatically generated reasons codes is not always appropriate.
1004 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
1005 using strerror. Previously, ERR_reason_error_string() returned
1006 library names as reason strings for SYSerr; but SYSerr is a special
1007 case where small numbers are errno values, not library numbers.
1010 *) Add '-dsaparam' option to 'openssl dhparam' application. This
1011 converts DSA parameters into DH parameters. (When creating parameters,
1012 DSA_generate_parameters is used.)
1015 *) Include 'length' (recommended exponent length) in C code generated
1016 by 'openssl dhparam -C'.
1019 *) The second argument to set_label in perlasm was already being used
1020 so couldn't be used as a "file scope" flag. Moved to third argument
1024 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
1025 instead of RAND_bytes for encryption IVs and salts.
1028 *) Include RAND_status() into RAND_METHOD instead of implementing
1029 it only for md_rand.c Otherwise replacing the PRNG by calling
1030 RAND_set_rand_method would be impossible.
1033 *) Don't let DSA_generate_key() enter an infinite loop if the random
1034 number generation fails.
1037 *) New 'rand' application for creating pseudo-random output.
1040 *) Added configuration support for Linux/IA64
1041 [Rolf Haberrecker <rolf@suse.de>]
1043 *) Assembler module support for Mingw32.
1046 *) Shared library support for HPUX (in shlib/).
1047 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
1049 *) Shared library support for Solaris gcc.
1050 [Lutz Behnke <behnke@trustcenter.de>]
1052 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
1054 *) PKCS7_encrypt() was adding text MIME headers twice because they
1055 were added manually and by SMIME_crlf_copy().
1058 *) In bntest.c don't call BN_rand with zero bits argument.
1059 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
1061 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
1062 case was implemented. This caused BN_div_recp() to fail occasionally.
1065 *) Add an optional second argument to the set_label() in the perl
1066 assembly language builder. If this argument exists and is set
1067 to 1 it signals that the assembler should use a symbol whose
1068 scope is the entire file, not just the current function. This
1069 is needed with MASM which uses the format label:: for this scope.
1070 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
1072 *) Change the ASN1 types so they are typedefs by default. Before
1073 almost all types were #define'd to ASN1_STRING which was causing
1074 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
1078 *) Change names of new functions to the new get1/get0 naming
1079 convention: After 'get1', the caller owns a reference count
1080 and has to call ..._free; 'get0' returns a pointer to some
1081 data structure without incrementing reference counters.
1082 (Some of the existing 'get' functions increment a reference
1083 counter, some don't.)
1084 Similarly, 'set1' and 'add1' functions increase reference
1085 counters or duplicate objects.
1088 *) Allow for the possibility of temp RSA key generation failure:
1089 the code used to assume it always worked and crashed on failure.
1092 *) Fix potential buffer overrun problem in BIO_printf().
1093 [Ulf Möller, using public domain code by Patrick Powell; problem
1094 pointed out by David Sacerdote <das33@cornell.edu>]
1096 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
1097 RAND_egd() and RAND_status(). In the command line application,
1098 the EGD socket can be specified like a seed file using RANDFILE
1102 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
1103 Some CAs (e.g. Verisign) distribute certificates in this form.
1106 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
1107 list to exclude them. This means that no special compilation option
1108 is needed to use anonymous DH: it just needs to be included in the
1112 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
1113 EVP_MD_type. The old functionality is available in a new macro called
1114 EVP_MD_md(). Change code that uses it and update docs.
1117 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
1118 where the 'void *' argument is replaced by a function pointer argument.
1119 Previously 'void *' was abused to point to functions, which works on
1120 many platforms, but is not correct. As these functions are usually
1121 called by macros defined in OpenSSL header files, most source code
1122 should work without changes.
1125 *) <openssl/opensslconf.h> (which is created by Configure) now contains
1126 sections with information on -D... compiler switches used for
1127 compiling the library so that applications can see them. To enable
1128 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
1129 must be defined. E.g.,
1130 #define OPENSSL_ALGORITHM_DEFINES
1131 #include <openssl/opensslconf.h>
1132 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
1133 [Richard Levitte, Ulf and Bodo Möller]
1135 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
1139 *) Change the 'other' type in certificate aux info to a STACK_OF
1140 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
1141 the required ASN1 format: arbitrary types determined by an OID.
1144 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
1145 argument to 'req'. This is not because the function is newer or
1146 better than others it just uses the work 'NEW' in the certificate
1147 request header lines. Some software needs this.
1150 *) Reorganise password command line arguments: now passwords can be
1151 obtained from various sources. Delete the PEM_cb function and make
1152 it the default behaviour: i.e. if the callback is NULL and the
1153 usrdata argument is not NULL interpret it as a null terminated pass
1154 phrase. If usrdata and the callback are NULL then the pass phrase
1155 is prompted for as usual.
1158 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
1159 the support is automatically enabled. The resulting binaries will
1160 autodetect the card and use it if present.
1161 [Ben Laurie and Compaq Inc.]
1163 *) Work around for Netscape hang bug. This sends certificate request
1164 and server done in one record. Since this is perfectly legal in the
1165 SSL/TLS protocol it isn't a "bug" option and is on by default. See
1166 the bugs/SSLv3 entry for more info.
1169 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
1172 *) Add -rand argument to smime and pkcs12 applications and read/write
1176 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
1179 *) Add command line password options to the remaining applications.
1182 *) Bug fix for BN_div_recp() for numerators with an even number of
1186 *) More tests in bntest.c, and changed test_bn output.
1189 *) ./config recognizes MacOS X now.
1192 *) Bug fix for BN_div() when the first words of num and divsor are
1193 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
1196 *) Add support for various broken PKCS#8 formats, and command line
1197 options to produce them.
1200 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
1201 get temporary BIGNUMs from a BN_CTX.
1204 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
1208 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
1209 include a #define from the old name to the new. The original intent
1210 was that statically linked binaries could for example just call
1211 SSLeay_add_all_ciphers() to just add ciphers to the table and not
1212 link with digests. This never worked becayse SSLeay_add_all_digests()
1213 and SSLeay_add_all_ciphers() were in the same source file so calling
1214 one would link with the other. They are now in separate source files.
1217 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
1220 *) Use a less unusual form of the Miller-Rabin primality test (it used
1221 a binary algorithm for exponentiation integrated into the Miller-Rabin
1222 loop, our standard modexp algorithms are faster).
1225 *) Support for the EBCDIC character set completed.
1226 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
1228 *) Source code cleanups: use const where appropriate, eliminate casts,
1229 use void * instead of char * in lhash.
1232 *) Bugfix: ssl3_send_server_key_exchange was not restartable
1233 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
1234 this the server could overwrite ephemeral keys that the client
1238 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
1239 using 50 iterations of the Rabin-Miller test.
1241 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
1242 iterations of the Rabin-Miller test as required by the appendix
1243 to FIPS PUB 186[-1]) instead of DSA_is_prime.
1244 As BN_is_prime_fasttest includes trial division, DSA parameter
1245 generation becomes much faster.
1247 This implies a change for the callback functions in DSA_is_prime
1248 and DSA_generate_parameters: The callback function is called once
1249 for each positive witness in the Rabin-Miller test, not just
1250 occasionally in the inner loop; and the parameters to the
1251 callback function now provide an iteration count for the outer
1252 loop rather than for the current invocation of the inner loop.
1253 DSA_generate_parameters additionally can call the callback
1254 function with an 'iteration count' of -1, meaning that a
1255 candidate has passed the trial division test (when q is generated
1256 from an application-provided seed, trial division is skipped).
1259 *) New function BN_is_prime_fasttest that optionally does trial
1260 division before starting the Rabin-Miller test and has
1261 an additional BN_CTX * argument (whereas BN_is_prime always
1262 has to allocate at least one BN_CTX).
1263 'callback(1, -1, cb_arg)' is called when a number has passed the
1264 trial division stage.
1267 *) Fix for bug in CRL encoding. The validity dates weren't being handled
1271 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
1274 *) New function BN_pseudo_rand().
1277 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
1278 bignum version of BN_from_montgomery() with the working code from
1279 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
1283 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
1284 made it impossible to use the same SSL_SESSION data structure in
1285 SSL2 clients in multiple threads.
1288 *) The return value of RAND_load_file() no longer counts bytes obtained
1289 by stat(). RAND_load_file(..., -1) is new and uses the complete file
1290 to seed the PRNG (previously an explicit byte count was required).
1291 [Ulf Möller, Bodo Möller]
1293 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
1294 used (char *) instead of (void *) and had casts all over the place.
1297 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
1300 *) Retain source code compatibility for BN_prime_checks macro:
1301 BN_is_prime(..., BN_prime_checks, ...) now uses
1302 BN_prime_checks_for_size to determine the appropriate number of
1303 Rabin-Miller iterations.
1306 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
1307 DH_CHECK_P_NOT_SAFE_PRIME.
1308 (Check if this is true? OpenPGP calls them "strong".)
1311 *) Merge the functionality of "dh" and "gendh" programs into a new program
1312 "dhparam". The old programs are retained for now but will handle DH keys
1313 (instead of parameters) in future.
1316 *) Make the ciphers, s_server and s_client programs check the return values
1317 when a new cipher list is set.
1320 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
1321 ciphers. Before when the 56bit ciphers were enabled the sorting was
1324 The syntax for the cipher sorting has been extended to support sorting by
1325 cipher-strength (using the strength_bits hard coded in the tables).
1326 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
1328 Fix a bug in the cipher-command parser: when supplying a cipher command
1329 string with an "undefined" symbol (neither command nor alphanumeric
1330 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
1331 an error is flagged.
1333 Due to the strength-sorting extension, the code of the
1334 ssl_create_cipher_list() function was completely rearranged. I hope that
1335 the readability was also increased :-)
1336 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
1338 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
1339 for the first serial number and places 2 in the serial number file. This
1340 avoids problems when the root CA is created with serial number zero and
1341 the first user certificate has the same issuer name and serial number
1345 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
1346 the new code. Add documentation for this stuff.
1349 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
1350 X509_*() to X509at_*() on the grounds that they don't handle X509
1351 structures and behave in an analagous way to the X509v3 functions:
1352 they shouldn't be called directly but wrapper functions should be used
1355 So we also now have some wrapper functions that call the X509at functions
1356 when passed certificate requests. (TO DO: similar things can be done with
1357 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
1358 things. Some of these need some d2i or i2d and print functionality
1359 because they handle more complex structures.)
1362 *) Add missing #ifndefs that caused missing symbols when building libssl
1363 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
1364 NO_RSA in ssl/s2*.c.
1365 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
1367 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
1368 has a return value which indicates the quality of the random data
1369 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
1370 error queue. New function RAND_pseudo_bytes() generates output that is
1371 guaranteed to be unique but not unpredictable. RAND_add is like
1372 RAND_seed, but takes an extra argument for an entropy estimate
1373 (RAND_seed always assumes full entropy).
1376 *) Do more iterations of Rabin-Miller probable prime test (specifically,
1377 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
1378 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
1379 in crypto/bn/bn_prime.c for the complete table). This guarantees a
1380 false-positive rate of at most 2^-80 for random input.
1383 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
1386 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
1387 in the 0.9.5 release), this returns the chain
1388 from an X509_CTX structure with a dup of the stack and all
1389 the X509 reference counts upped: so the stack will exist
1390 after X509_CTX_cleanup() has been called. Modify pkcs12.c
1393 Also make SSL_SESSION_print() print out the verify return
1397 *) Add manpage for the pkcs12 command. Also change the default
1398 behaviour so MAC iteration counts are used unless the new
1399 -nomaciter option is used. This improves file security and
1400 only older versions of MSIE (4.0 for example) need it.
1403 *) Honor the no-xxx Configure options when creating .DEF files.
1406 *) Add PKCS#10 attributes to field table: challengePassword,
1407 unstructuredName and unstructuredAddress. These are taken from
1408 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
1409 international characters are used.
1411 More changes to X509_ATTRIBUTE code: allow the setting of types
1412 based on strings. Remove the 'loc' parameter when adding
1413 attributes because these will be a SET OF encoding which is sorted
1417 *) Initial changes to the 'req' utility to allow request generation
1418 automation. This will allow an application to just generate a template
1419 file containing all the field values and have req construct the
1422 Initial support for X509_ATTRIBUTE handling. Stacks of these are
1423 used all over the place including certificate requests and PKCS#7
1424 structures. They are currently handled manually where necessary with
1425 some primitive wrappers for PKCS#7. The new functions behave in a
1426 manner analogous to the X509 extension functions: they allow
1427 attributes to be looked up by NID and added.
1429 Later something similar to the X509V3 code would be desirable to
1430 automatically handle the encoding, decoding and printing of the
1431 more complex types. The string types like challengePassword can
1432 be handled by the string table functions.
1434 Also modified the multi byte string table handling. Now there is
1435 a 'global mask' which masks out certain types. The table itself
1436 can use the flag STABLE_NO_MASK to ignore the mask setting: this
1437 is useful when for example there is only one permissible type
1438 (as in countryName) and using the mask might result in no valid
1442 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
1443 SSL_get_peer_finished to allow applications to obtain the latest
1444 Finished messages sent to the peer or expected from the peer,
1445 respectively. (SSL_get_peer_finished is usually the Finished message
1446 actually received from the peer, otherwise the protocol will be aborted.)
1448 As the Finished message are message digests of the complete handshake
1449 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
1450 be used for external authentication procedures when the authentication
1451 provided by SSL/TLS is not desired or is not enough.
1454 *) Enhanced support for Alpha Linux is added. Now ./config checks if
1455 the host supports BWX extension and if Compaq C is present on the
1456 $PATH. Just exploiting of the BWX extension results in 20-30%
1457 performance kick for some algorithms, e.g. DES and RC4 to mention
1458 a couple. Compaq C in turn generates ~20% faster code for MD5 and
1462 *) Add support for MS "fast SGC". This is arguably a violation of the
1463 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
1464 weak crypto and after checking the certificate is SGC a second one
1465 with strong crypto. MS SGC stops the first handshake after receiving
1466 the server certificate message and sends a second client hello. Since
1467 a server will typically do all the time consuming operations before
1468 expecting any further messages from the client (server key exchange
1469 is the most expensive) there is little difference between the two.
1471 To get OpenSSL to support MS SGC we have to permit a second client
1472 hello message after we have sent server done. In addition we have to
1473 reset the MAC if we do get this second client hello.
1476 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
1477 if a DER encoded private key is RSA or DSA traditional format. Changed
1478 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
1479 format DER encoded private key. Newer code should use PKCS#8 format which
1480 has the key type encoded in the ASN1 structure. Added DER private key
1481 support to pkcs8 application.
1484 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
1485 ciphersuites has been selected (as required by the SSL 3/TLS 1
1486 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
1487 is set, we interpret this as a request to violate the specification
1488 (the worst that can happen is a handshake failure, and 'correct'
1489 behaviour would result in a handshake failure anyway).
1492 *) In SSL_CTX_add_session, take into account that there might be multiple
1493 SSL_SESSION structures with the same session ID (e.g. when two threads
1494 concurrently obtain them from an external cache).
1495 The internal cache can handle only one SSL_SESSION with a given ID,
1496 so if there's a conflict, we now throw out the old one to achieve
1500 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
1501 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
1502 some routines that use cipher OIDs: some ciphers do not have OIDs
1503 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
1507 *) Simplify the trust setting structure and code. Now we just have
1508 two sequences of OIDs for trusted and rejected settings. These will
1509 typically have values the same as the extended key usage extension
1510 and any application specific purposes.
1512 The trust checking code now has a default behaviour: it will just
1513 check for an object with the same NID as the passed id. Functions can
1514 be provided to override either the default behaviour or the behaviour
1515 for a given id. SSL client, server and email already have functions
1516 in place for compatibility: they check the NID and also return "trusted"
1517 if the certificate is self signed.
1520 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
1521 traditional format into an EVP_PKEY structure.
1524 *) Add a password callback function PEM_cb() which either prompts for
1525 a password if usr_data is NULL or otherwise assumes it is a null
1526 terminated password. Allow passwords to be passed on command line
1527 environment or config files in a few more utilities.
1530 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
1531 keys. Add some short names for PKCS#8 PBE algorithms and allow them
1532 to be specified on the command line for the pkcs8 and pkcs12 utilities.
1533 Update documentation.
1536 *) Support for ASN1 "NULL" type. This could be handled before by using
1537 ASN1_TYPE but there wasn't any function that would try to read a NULL
1538 and produce an error if it couldn't. For compatibility we also have
1539 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
1540 don't allocate anything because they don't need to.
1543 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
1545 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
1547 *) Rebuild of the memory allocation routines used by OpenSSL code and
1548 possibly others as well. The purpose is to make an interface that
1549 provide hooks so anyone can build a separate set of allocation and
1550 deallocation routines to be used by OpenSSL, for example memory
1551 pool implementations, or something else, which was previously hard
1552 since Malloc(), Realloc() and Free() were defined as macros having
1553 the values malloc, realloc and free, respectively (except for Win32
1554 compilations). The same is provided for memory debugging code.
1555 OpenSSL already comes with functionality to find memory leaks, but
1556 this gives people a chance to debug other memory problems.
1558 With these changes, a new set of functions and macros have appeared:
1560 CRYPTO_set_mem_debug_functions() [F]
1561 CRYPTO_get_mem_debug_functions() [F]
1562 CRYPTO_dbg_set_options() [F]
1563 CRYPTO_dbg_get_options() [F]
1564 CRYPTO_malloc_debug_init() [M]
1566 The memory debug functions are NULL by default, unless the library
1567 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
1568 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
1569 gives the standard debugging functions that come with OpenSSL) or
1570 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
1571 provided by the library user) must be used. When the standard
1572 debugging functions are used, CRYPTO_dbg_set_options can be used to
1573 request additional information:
1574 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
1575 the CRYPTO_MDEBUG_xxx macro when compiling the library.
1577 Also, things like CRYPTO_set_mem_functions will always give the
1578 expected result (the new set of functions is used for allocation
1579 and deallocation) at all times, regardless of platform and compiler
1582 To finish it up, some functions that were never use in any other
1583 way than through macros have a new API and new semantic:
1586 CRYPTO_dbg_realloc()
1589 All macros of value have retained their old syntax.
1590 [Richard Levitte and Bodo Moeller]
1592 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
1593 ordering of SMIMECapabilities wasn't in "strength order" and there
1594 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
1598 *) Some ASN1 types with illegal zero length encoding (INTEGER,
1599 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
1600 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
1602 *) Merge in my S/MIME library for OpenSSL. This provides a simple
1603 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
1604 functionality to handle multipart/signed properly) and a utility
1605 called 'smime' to call all this stuff. This is based on code I
1606 originally wrote for Celo who have kindly allowed it to be
1607 included in OpenSSL.
1610 *) Add variants des_set_key_checked and des_set_key_unchecked of
1611 des_set_key (aka des_key_sched). Global variable des_check_key
1612 decides which of these is called by des_set_key; this way
1613 des_check_key behaves as it always did, but applications and
1614 the library itself, which was buggy for des_check_key == 1,
1615 have a cleaner way to pick the version they need.
1618 *) New function PKCS12_newpass() which changes the password of a
1622 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
1623 dynamic mix. In both cases the ids can be used as an index into the
1624 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
1625 functions so they accept a list of the field values and the
1626 application doesn't need to directly manipulate the X509_TRUST
1630 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
1634 *) Modify the way the V3 extension code looks up extensions. This now
1635 works in a similar way to the object code: we have some "standard"
1636 extensions in a static table which is searched with OBJ_bsearch()
1637 and the application can add dynamic ones if needed. The file
1638 crypto/x509v3/ext_dat.h now has the info: this file needs to be
1639 updated whenever a new extension is added to the core code and kept
1640 in ext_nid order. There is a simple program 'tabtest.c' which checks
1641 this. New extensions are not added too often so this file can readily
1642 be maintained manually.
1644 There are two big advantages in doing things this way. The extensions
1645 can be looked up immediately and no longer need to be "added" using
1646 X509V3_add_standard_extensions(): this function now does nothing.
1647 [Side note: I get *lots* of email saying the extension code doesn't
1648 work because people forget to call this function]
1649 Also no dynamic allocation is done unless new extensions are added:
1650 so if we don't add custom extensions there is no need to call
1651 X509V3_EXT_cleanup().
1654 *) Modify enc utility's salting as follows: make salting the default. Add a
1655 magic header, so unsalted files fail gracefully instead of just decrypting
1656 to garbage. This is because not salting is a big security hole, so people
1657 should be discouraged from doing it.
1660 *) Fixes and enhancements to the 'x509' utility. It allowed a message
1661 digest to be passed on the command line but it only used this
1662 parameter when signing a certificate. Modified so all relevant
1663 operations are affected by the digest parameter including the
1664 -fingerprint and -x509toreq options. Also -x509toreq choked if a
1665 DSA key was used because it didn't fix the digest.
1668 *) Initial certificate chain verify code. Currently tests the untrusted
1669 certificates for consistency with the verify purpose (which is set
1670 when the X509_STORE_CTX structure is set up) and checks the pathlength.
1672 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
1673 this is because it will reject chains with invalid extensions whereas
1674 every previous version of OpenSSL and SSLeay made no checks at all.
1676 Trust code: checks the root CA for the relevant trust settings. Trust
1677 settings have an initial value consistent with the verify purpose: e.g.
1678 if the verify purpose is for SSL client use it expects the CA to be
1679 trusted for SSL client use. However the default value can be changed to
1680 permit custom trust settings: one example of this would be to only trust
1681 certificates from a specific "secure" set of CAs.
1683 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
1684 which should be used for version portability: especially since the
1685 verify structure is likely to change more often now.
1687 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
1688 to set them. If not set then assume SSL clients will verify SSL servers
1691 Two new options to the verify program: -untrusted allows a set of
1692 untrusted certificates to be passed in and -purpose which sets the
1693 intended purpose of the certificate. If a purpose is set then the
1694 new chain verify code is used to check extension consistency.
1697 *) Support for the authority information access extension.
1700 *) Modify RSA and DSA PEM read routines to transparently handle
1701 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
1702 public keys in a format compatible with certificate
1703 SubjectPublicKeyInfo structures. Unfortunately there were already
1704 functions called *_PublicKey_* which used various odd formats so
1705 these are retained for compatibility: however the DSA variants were
1706 never in a public release so they have been deleted. Changed dsa/rsa
1707 utilities to handle the new format: note no releases ever handled public
1708 keys so we should be OK.
1710 The primary motivation for this change is to avoid the same fiasco
1711 that dogs private keys: there are several incompatible private key
1712 formats some of which are standard and some OpenSSL specific and
1713 require various evil hacks to allow partial transparent handling and
1714 even then it doesn't work with DER formats. Given the option anything
1715 other than PKCS#8 should be dumped: but the other formats have to
1716 stay in the name of compatibility.
1718 With public keys and the benefit of hindsight one standard format
1719 is used which works with EVP_PKEY, RSA or DSA structures: though
1720 it clearly returns an error if you try to read the wrong kind of key.
1722 Added a -pubkey option to the 'x509' utility to output the public key.
1723 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
1724 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
1725 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
1726 that do the same as the EVP_PKEY_assign_*() except they up the
1727 reference count of the added key (they don't "swallow" the
1731 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
1732 CRLs would fail if the file contained no certificates or no CRLs:
1733 added a new function to read in both types and return the number
1734 read: this means that if none are read it will be an error. The
1735 DER versions of the certificate and CRL reader would always fail
1736 because it isn't possible to mix certificates and CRLs in DER format
1737 without choking one or the other routine. Changed this to just read
1738 a certificate: this is the best we can do. Also modified the code
1739 in apps/verify.c to take notice of return codes: it was previously
1740 attempting to read in certificates from NULL pointers and ignoring
1741 any errors: this is one reason why the cert and CRL reader seemed
1742 to work. It doesn't check return codes from the default certificate
1743 routines: these may well fail if the certificates aren't installed.
1746 *) Code to support otherName option in GeneralName.
1749 *) First update to verify code. Change the verify utility
1750 so it warns if it is passed a self signed certificate:
1751 for consistency with the normal behaviour. X509_verify
1752 has been modified to it will now verify a self signed
1753 certificate if *exactly* the same certificate appears
1754 in the store: it was previously impossible to trust a
1755 single self signed certificate. This means that:
1756 openssl verify ss.pem
1757 now gives a warning about a self signed certificate but
1758 openssl verify -CAfile ss.pem ss.pem
1762 *) For servers, store verify_result in SSL_SESSION data structure
1763 (and add it to external session representation).
1764 This is needed when client certificate verifications fails,
1765 but an application-provided verification callback (set by
1766 SSL_CTX_set_cert_verify_callback) allows accepting the session
1767 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
1768 but returns 1): When the session is reused, we have to set
1769 ssl->verify_result to the appropriate error code to avoid
1771 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
1773 *) Fix a bug in the new PKCS#7 code: it didn't consider the
1774 case in PKCS7_dataInit() where the signed PKCS7 structure
1775 didn't contain any existing data because it was being created.
1776 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
1778 *) Add a salt to the key derivation routines in enc.c. This
1779 forms the first 8 bytes of the encrypted file. Also add a
1780 -S option to allow a salt to be input on the command line.
1783 *) New function X509_cmp(). Oddly enough there wasn't a function
1784 to compare two certificates. We do this by working out the SHA1
1785 hash and comparing that. X509_cmp() will be needed by the trust
1789 *) SSL_get1_session() is like SSL_get_session(), but increments
1790 the reference count in the SSL_SESSION returned.
1791 [Geoff Thorpe <geoff@eu.c2.net>]
1793 *) Fix for 'req': it was adding a null to request attributes.
1794 Also change the X509_LOOKUP and X509_INFO code to handle
1795 certificate auxiliary information.
1798 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
1802 *) Add the possibility to add extra information to the memory leak
1803 detecting output, to form tracebacks, showing from where each
1804 allocation was originated: CRYPTO_push_info("constant string") adds
1805 the string plus current file name and line number to a per-thread
1806 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
1807 is like calling CYRPTO_pop_info() until the stack is empty.
1808 Also updated memory leak detection code to be multi-thread-safe.
1811 *) Add options -text and -noout to pkcs7 utility and delete the
1812 encryption options which never did anything. Update docs.
1815 *) Add options to some of the utilities to allow the pass phrase
1816 to be included on either the command line (not recommended on
1817 OSes like Unix) or read from the environment. Update the
1818 manpages and fix a few bugs.
1821 *) Add a few manpages for some of the openssl commands.
1824 *) Fix the -revoke option in ca. It was freeing up memory twice,
1825 leaking and not finding already revoked certificates.
1828 *) Extensive changes to support certificate auxiliary information.
1829 This involves the use of X509_CERT_AUX structure and X509_AUX
1830 functions. An X509_AUX function such as PEM_read_X509_AUX()
1831 can still read in a certificate file in the usual way but it
1832 will also read in any additional "auxiliary information". By
1833 doing things this way a fair degree of compatibility can be
1834 retained: existing certificates can have this information added
1835 using the new 'x509' options.
1837 Current auxiliary information includes an "alias" and some trust
1838 settings. The trust settings will ultimately be used in enhanced
1839 certificate chain verification routines: currently a certificate
1840 can only be trusted if it is self signed and then it is trusted
1844 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
1845 The problem was that one of the replacement routines had not been working
1846 since SSLeay releases. For now the offending routine has been replaced
1847 with non-optimised assembler. Even so, this now gives around 95%
1848 performance improvement for 1024 bit RSA signs.
1851 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
1852 handling. Most clients have the effective key size in bits equal to
1853 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
1854 A few however don't do this and instead use the size of the decrypted key
1855 to determine the RC2 key length and the AlgorithmIdentifier to determine
1856 the effective key length. In this case the effective key length can still
1857 be 40 bits but the key length can be 168 bits for example. This is fixed
1858 by manually forcing an RC2 key into the EVP_PKEY structure because the
1859 EVP code can't currently handle unusual RC2 key sizes: it always assumes
1860 the key length and effective key length are equal.
1863 *) Add a bunch of functions that should simplify the creation of
1864 X509_NAME structures. Now you should be able to do:
1865 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
1866 and have it automatically work out the correct field type and fill in
1867 the structures. The more adventurous can try:
1868 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
1869 and it will (hopefully) work out the correct multibyte encoding.
1872 *) Change the 'req' utility to use the new field handling and multibyte
1873 copy routines. Before the DN field creation was handled in an ad hoc
1874 way in req, ca, and x509 which was rather broken and didn't support
1875 BMPStrings or UTF8Strings. Since some software doesn't implement
1876 BMPStrings or UTF8Strings yet, they can be enabled using the config file
1877 using the dirstring_type option. See the new comment in the default
1878 openssl.cnf for more info.
1881 *) Make crypto/rand/md_rand.c more robust:
1882 - Assure unique random numbers after fork().
1883 - Make sure that concurrent threads access the global counter and
1884 md serializably so that we never lose entropy in them
1885 or use exactly the same state in multiple threads.
1886 Access to the large state is not always serializable because
1887 the additional locking could be a performance killer, and
1888 md should be large enough anyway.
1891 *) New file apps/app_rand.c with commonly needed functionality
1892 for handling the random seed file.
1894 Use the random seed file in some applications that previously did not:
1896 dsaparam -genkey (which also ignored its '-rand' option),
1899 x509 (when signing).
1900 Except on systems with /dev/urandom, it is crucial to have a random
1901 seed file at least for key creation, DSA signing, and for DH exchanges;
1902 for RSA signatures we could do without one.
1904 gendh and gendsa (unlike genrsa) used to read only the first byte
1905 of each file listed in the '-rand' option. The function as previously
1906 found in genrsa is now in app_rand.c and is used by all programs
1907 that support '-rand'.
1910 *) In RAND_write_file, use mode 0600 for creating files;
1911 don't just chmod when it may be too late.
1914 *) Report an error from X509_STORE_load_locations
1915 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1918 *) New function ASN1_mbstring_copy() this copies a string in either
1919 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1920 into an ASN1_STRING type. A mask of permissible types is passed
1921 and it chooses the "minimal" type to use or an error if not type
1925 *) Add function equivalents to the various macros in asn1.h. The old
1926 macros are retained with an M_ prefix. Code inside the library can
1927 use the M_ macros. External code (including the openssl utility)
1928 should *NOT* in order to be "shared library friendly".
1931 *) Add various functions that can check a certificate's extensions
1932 to see if it usable for various purposes such as SSL client,
1933 server or S/MIME and CAs of these types. This is currently
1934 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1935 verification. Also added a -purpose flag to x509 utility to
1936 print out all the purposes.
1939 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1943 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1944 for, obtain and decode and extension and obtain its critical flag.
1945 This allows all the necessary extension code to be handled in a
1946 single function call.
1949 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1950 platforms. See crypto/rc4/rc4_enc.c for further details.
1953 *) New -noout option to asn1parse. This causes no output to be produced
1954 its main use is when combined with -strparse and -out to extract data
1955 from a file (which may not be in ASN.1 format).
1958 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1959 when producing the local key id.
1960 [Richard Levitte <levitte@stacken.kth.se>]
1962 *) New option -dhparam in s_server. This allows a DH parameter file to be
1963 stated explicitly. If it is not stated then it tries the first server
1964 certificate file. The previous behaviour hard coded the filename
1968 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1969 a public key to be input or output. For example:
1970 openssl rsa -in key.pem -pubout -out pubkey.pem
1971 Also added necessary DSA public key functions to handle this.
1974 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1975 in the message. This was handled by allowing
1976 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1977 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1979 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1980 to the end of the strings whereas this didn't. This would cause problems
1981 if strings read with d2i_ASN1_bytes() were later modified.
1982 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1984 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1985 data and it contains EOF it will end up returning an error. This is
1986 caused by input 46 bytes long. The cause is due to the way base64
1987 BIOs find the start of base64 encoded data. They do this by trying a
1988 trial decode on each line until they find one that works. When they
1989 do a flag is set and it starts again knowing it can pass all the
1990 data directly through the decoder. Unfortunately it doesn't reset
1991 the context it uses. This means that if EOF is reached an attempt
1992 is made to pass two EOFs through the context and this causes the
1993 resulting error. This can also cause other problems as well. As is
1994 usual with these problems it takes *ages* to find and the fix is
1995 trivial: move one line.
1996 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1998 *) Ugly workaround to get s_client and s_server working under Windows. The
1999 old code wouldn't work because it needed to select() on sockets and the
2000 tty (for keypresses and to see if data could be written). Win32 only
2001 supports select() on sockets so we select() with a 1s timeout on the
2002 sockets and then see if any characters are waiting to be read, if none
2003 are present then we retry, we also assume we can always write data to
2004 the tty. This isn't nice because the code then blocks until we've
2005 received a complete line of data and it is effectively polling the
2006 keyboard at 1s intervals: however it's quite a bit better than not
2007 working at all :-) A dedicated Windows application might handle this
2008 with an event loop for example.
2011 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
2012 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
2013 will be called when RSA_sign() and RSA_verify() are used. This is useful
2014 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
2015 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
2016 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
2017 This necessitated the support of an extra signature type NID_md5_sha1
2018 for SSL signatures and modifications to the SSL library to use it instead
2019 of calling RSA_public_decrypt() and RSA_private_encrypt().
2022 *) Add new -verify -CAfile and -CApath options to the crl program, these
2023 will lookup a CRL issuers certificate and verify the signature in a
2024 similar way to the verify program. Tidy up the crl program so it
2025 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
2026 less strict. It will now permit CRL extensions even if it is not
2027 a V2 CRL: this will allow it to tolerate some broken CRLs.
2030 *) Initialize all non-automatic variables each time one of the openssl
2031 sub-programs is started (this is necessary as they may be started
2032 multiple times from the "OpenSSL>" prompt).
2033 [Lennart Bang, Bodo Moeller]
2035 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
2036 removing all other RSA functionality (this is what NO_RSA does). This
2037 is so (for example) those in the US can disable those operations covered
2038 by the RSA patent while allowing storage and parsing of RSA keys and RSA
2042 *) Non-copying interface to BIO pairs.
2043 (still largely untested)
2046 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
2047 ASCII string. This was handled independently in various places before.
2050 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
2051 UTF8 strings a character at a time.
2054 *) Use client_version from client hello to select the protocol
2055 (s23_srvr.c) and for RSA client key exchange verification
2056 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
2059 *) Add various utility functions to handle SPKACs, these were previously
2060 handled by poking round in the structure internals. Added new function
2061 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
2062 print, verify and generate SPKACs. Based on an original idea from
2063 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
2066 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
2069 *) Allow the config file extension section to be overwritten on the
2070 command line. Based on an original idea from Massimiliano Pala
2071 <madwolf@comune.modena.it>. The new option is called -extensions
2072 and can be applied to ca, req and x509. Also -reqexts to override
2073 the request extensions in req and -crlexts to override the crl extensions
2077 *) Add new feature to the SPKAC handling in ca. Now you can include
2078 the same field multiple times by preceding it by "XXXX." for example:
2081 this is the same syntax as used in the req config file.
2084 *) Allow certificate extensions to be added to certificate requests. These
2085 are specified in a 'req_extensions' option of the req section of the
2086 config file. They can be printed out with the -text option to req but
2087 are otherwise ignored at present.
2090 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
2091 data read consists of only the final block it would not decrypted because
2092 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
2093 A misplaced 'break' also meant the decrypted final block might not be
2094 copied until the next read.
2097 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
2098 a few extra parameters to the DH structure: these will be useful if
2099 for example we want the value of 'q' or implement X9.42 DH.
2102 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
2103 provides hooks that allow the default DSA functions or functions on a
2104 "per key" basis to be replaced. This allows hardware acceleration and
2105 hardware key storage to be handled without major modification to the
2106 library. Also added low level modexp hooks and CRYPTO_EX structure and
2107 associated functions.
2110 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
2111 as "read only": it can't be written to and the buffer it points to will
2112 not be freed. Reading from a read only BIO is much more efficient than
2113 a normal memory BIO. This was added because there are several times when
2114 an area of memory needs to be read from a BIO. The previous method was
2115 to create a memory BIO and write the data to it, this results in two
2116 copies of the data and an O(n^2) reading algorithm. There is a new
2117 function BIO_new_mem_buf() which creates a read only memory BIO from
2118 an area of memory. Also modified the PKCS#7 routines to use read only
2122 *) Bugfix: ssl23_get_client_hello did not work properly when called in
2123 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
2124 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
2125 but a retry condition occured while trying to read the rest.
2128 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
2129 NID_pkcs7_encrypted by default: this was wrong since this should almost
2130 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
2131 the encrypted data type: this is a more sensible place to put it and it
2132 allows the PKCS#12 code to be tidied up that duplicated this
2136 *) Changed obj_dat.pl script so it takes its input and output files on
2137 the command line. This should avoid shell escape redirection problems
2141 *) Initial support for certificate extension requests, these are included
2142 in things like Xenroll certificate requests. Included functions to allow
2143 extensions to be obtained and added.
2146 *) -crlf option to s_client and s_server for sending newlines as
2147 CRLF (as required by many protocols).
2150 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
2152 *) Install libRSAglue.a when OpenSSL is built with RSAref.
2153 [Ralf S. Engelschall]
2155 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
2156 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
2158 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
2162 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
2163 DH parameters/keys (q is lost during that conversion, but the resulting
2164 DH parameters contain its length).
2166 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
2167 much faster than DH_generate_parameters (which creates parameters
2168 where p = 2*q + 1), and also the smaller q makes DH computations
2169 much more efficient (160-bit exponentiation instead of 1024-bit
2170 exponentiation); so this provides a convenient way to support DHE
2171 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
2172 utter importance to use
2173 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
2175 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
2176 when such DH parameters are used, because otherwise small subgroup
2177 attacks may become possible!
2180 *) Avoid memory leak in i2d_DHparams.
2183 *) Allow the -k option to be used more than once in the enc program:
2184 this allows the same encrypted message to be read by multiple recipients.
2187 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
2188 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
2189 it will always use the numerical form of the OID, even if it has a short
2193 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
2194 method only got called if p,q,dmp1,dmq1,iqmp components were present,
2195 otherwise bn_mod_exp was called. In the case of hardware keys for example
2196 no private key components need be present and it might store extra data
2197 in the RSA structure, which cannot be accessed from bn_mod_exp.
2198 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
2199 private key operations.
2202 *) Added support for SPARC Linux.
2205 *) pem_password_cb function type incompatibly changed from
2206 typedef int pem_password_cb(char *buf, int size, int rwflag);
2208 ....(char *buf, int size, int rwflag, void *userdata);
2209 so that applications can pass data to their callbacks:
2210 The PEM[_ASN1]_{read,write}... functions and macros now take an
2211 additional void * argument, which is just handed through whenever
2212 the password callback is called.
2213 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
2215 New function SSL_CTX_set_default_passwd_cb_userdata.
2217 Compatibility note: As many C implementations push function arguments
2218 onto the stack in reverse order, the new library version is likely to
2219 interoperate with programs that have been compiled with the old
2220 pem_password_cb definition (PEM_whatever takes some data that
2221 happens to be on the stack as its last argument, and the callback
2222 just ignores this garbage); but there is no guarantee whatsoever that
2225 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
2226 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
2227 problems not only on Windows, but also on some Unix platforms.
2228 To avoid problematic command lines, these definitions are now in an
2229 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
2230 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
2233 *) MIPS III/IV assembler module is reimplemented.
2236 *) More DES library cleanups: remove references to srand/rand and
2237 delete an unused file.
2240 *) Add support for the the free Netwide assembler (NASM) under Win32,
2241 since not many people have MASM (ml) and it can be hard to obtain.
2242 This is currently experimental but it seems to work OK and pass all
2243 the tests. Check out INSTALL.W32 for info.
2246 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
2247 without temporary keys kept an extra copy of the server key,
2248 and connections with temporary keys did not free everything in case
2252 *) New function RSA_check_key and new openssl rsa option -check
2253 for verifying the consistency of RSA keys.
2254 [Ulf Moeller, Bodo Moeller]
2256 *) Various changes to make Win32 compile work:
2257 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
2258 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
2259 comparison" warnings.
2260 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
2263 *) Add a debugging option to PKCS#5 v2 key generation function: when
2264 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
2265 derived keys are printed to stderr.
2268 *) Copy the flags in ASN1_STRING_dup().
2269 [Roman E. Pavlov <pre@mo.msk.ru>]
2271 *) The x509 application mishandled signing requests containing DSA
2272 keys when the signing key was also DSA and the parameters didn't match.
2274 It was supposed to omit the parameters when they matched the signing key:
2275 the verifying software was then supposed to automatically use the CA's
2276 parameters if they were absent from the end user certificate.
2278 Omitting parameters is no longer recommended. The test was also
2279 the wrong way round! This was probably due to unusual behaviour in
2280 EVP_cmp_parameters() which returns 1 if the parameters match.
2281 This meant that parameters were omitted when they *didn't* match and
2282 the certificate was useless. Certificates signed with 'ca' didn't have
2284 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
2286 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
2287 The interface is as follows:
2288 Applications can use
2289 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
2290 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
2291 "off" is now the default.
2292 The library internally uses
2293 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
2294 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
2295 to disable memory-checking temporarily.
2297 Some inconsistent states that previously were possible (and were
2298 even the default) are now avoided.
2300 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
2301 with each memory chunk allocated; this is occasionally more helpful
2302 than just having a counter.
2304 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
2306 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
2310 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
2311 which largely parallels "options", but is for changing API behaviour,
2312 whereas "options" are about protocol behaviour.
2313 Initial "mode" flags are:
2315 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
2316 a single record has been written.
2317 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
2318 retries use the same buffer location.
2319 (But all of the contents must be
2323 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
2326 *) Fix problems with no-hmac etc.
2327 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
2329 *) New functions RSA_get_default_method(), RSA_set_method() and
2330 RSA_get_method(). These allows replacement of RSA_METHODs without having
2331 to mess around with the internals of an RSA structure.
2334 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
2335 Also really enable memory leak checks in openssl.c and in some
2337 [Chad C. Mulligan, Bodo Moeller]
2339 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
2340 up the length of negative integers. This has now been simplified to just
2341 store the length when it is first determined and use it later, rather
2342 than trying to keep track of where data is copied and updating it to
2344 [Steve Henson, reported by Brien Wheeler
2345 <bwheeler@authentica-security.com>]
2347 *) Add a new function PKCS7_signatureVerify. This allows the verification
2348 of a PKCS#7 signature but with the signing certificate passed to the
2349 function itself. This contrasts with PKCS7_dataVerify which assumes the
2350 certificate is present in the PKCS#7 structure. This isn't always the
2351 case: certificates can be omitted from a PKCS#7 structure and be
2352 distributed by "out of band" means (such as a certificate database).
2355 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
2356 function prototypes in pem.h, also change util/mkdef.pl to add the
2357 necessary function names.
2360 *) mk1mf.pl (used by Windows builds) did not properly read the
2361 options set by Configure in the top level Makefile, and Configure
2362 was not even able to write more than one option correctly.
2363 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
2366 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
2367 file to be loaded from a BIO or FILE pointer. The BIO version will
2368 for example allow memory BIOs to contain config info.
2371 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
2372 Whoever hopes to achieve shared-library compatibility across versions
2373 must use this, not the compile-time macro.
2374 (Exercise 0.9.4: Which is the minimum library version required by
2376 Note: All this applies only to multi-threaded programs, others don't
2380 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
2381 through a BIO pair triggered the default case, i.e.
2382 SSLerr(...,SSL_R_UNKNOWN_STATE).
2385 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
2386 can use the SSL library even if none of the specific BIOs is
2390 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
2391 for the encoded length.
2392 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
2394 *) Add initial documentation of the X509V3 functions.
2397 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
2398 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
2399 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
2400 secure PKCS#8 private key format with a high iteration count.
2403 *) Fix determination of Perl interpreter: A perl or perl5
2404 _directory_ in $PATH was also accepted as the interpreter.
2405 [Ralf S. Engelschall]
2407 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
2408 wrong with it but it was very old and did things like calling
2409 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
2413 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
2414 to use the new extension code.
2417 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
2418 with macros. This should make it easier to change their form, add extra
2419 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
2423 *) Add to configuration table a new entry that can specify an alternative
2424 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
2425 according to Mark Crispin <MRC@Panda.COM>.
2429 *) DES CBC did not update the IV. Weird.
2432 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
2433 Changing the behaviour of the former might break existing programs --
2434 where IV updating is needed, des_ncbc_encrypt can be used.
2437 *) When bntest is run from "make test" it drives bc to check its
2438 calculations, as well as internally checking them. If an internal check
2439 fails, it needs to cause bc to give a non-zero result or make test carries
2440 on without noticing the failure. Fixed.
2443 *) DES library cleanups.
2446 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
2447 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
2448 ciphers. NOTE: although the key derivation function has been verified
2449 against some published test vectors it has not been extensively tested
2450 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
2454 *) Instead of "mkdir -p", which is not fully portable, use new
2455 Perl script "util/mkdir-p.pl".
2458 *) Rewrite the way password based encryption (PBE) is handled. It used to
2459 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
2460 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
2461 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
2462 the 'parameter' field of the AlgorithmIdentifier is passed to the
2463 underlying key generation function so it must do its own ASN1 parsing.
2464 This has also changed the EVP_PBE_CipherInit() function which now has a
2465 'parameter' argument instead of literal salt and iteration count values
2466 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
2469 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
2470 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
2471 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
2472 KEY" because this clashed with PKCS#8 unencrypted string. Since this
2473 value was just used as a "magic string" and not used directly its
2474 value doesn't matter.
2477 *) Introduce some semblance of const correctness to BN. Shame C doesn't
2481 *) "linux-sparc64" configuration (ultrapenguin).
2482 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
2483 "linux-sparc" configuration.
2484 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
2486 *) config now generates no-xxx options for missing ciphers.
2489 *) Support the EBCDIC character set (work in progress).
2490 File ebcdic.c not yet included because it has a different license.
2491 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2493 *) Support BS2000/OSD-POSIX.
2494 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2496 *) Make callbacks for key generation use void * instead of char *.
2499 *) Make S/MIME samples compile (not yet tested).
2502 *) Additional typesafe stacks.
2505 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
2509 Changes between 0.9.3 and 0.9.3a [29 May 1999]
2511 *) New configuration variant "sco5-gcc".
2513 *) Updated some demos.
2514 [Sean O Riordain, Wade Scholine]
2516 *) Add missing BIO_free at exit of pkcs12 application.
2519 *) Fix memory leak in conf.c.
2522 *) Updates for Win32 to assembler version of MD5.
2525 *) Set #! path to perl in apps/der_chop to where we found it
2526 instead of using a fixed path.
2529 *) SHA library changes for irix64-mips4-cc.
2532 *) Improvements for VMS support.
2536 Changes between 0.9.2b and 0.9.3 [24 May 1999]
2538 *) Bignum library bug fix. IRIX 6 passes "make test" now!
2539 This also avoids the problems with SC4.2 and unpatched SC5.
2540 [Andy Polyakov <appro@fy.chalmers.se>]
2542 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
2543 These are required because of the typesafe stack would otherwise break
2544 existing code. If old code used a structure member which used to be STACK
2545 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
2546 sk_num or sk_value it would produce an error because the num, data members
2547 are not present in STACK_OF. Now it just produces a warning. sk_set
2548 replaces the old method of assigning a value to sk_value
2549 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
2550 that does this will no longer work (and should use sk_set instead) but
2551 this could be regarded as a "questionable" behaviour anyway.
2554 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
2555 correctly handle encrypted S/MIME data.
2558 *) Change type of various DES function arguments from des_cblock
2559 (which means, in function argument declarations, pointer to char)
2560 to des_cblock * (meaning pointer to array with 8 char elements),
2561 which allows the compiler to do more typechecking; it was like
2562 that back in SSLeay, but with lots of ugly casts.
2564 Introduce new type const_des_cblock.
2567 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
2568 problems: find RecipientInfo structure that matches recipient certificate
2569 and initialise the ASN1 structures properly based on passed cipher.
2572 *) Belatedly make the BN tests actually check the results.
2575 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
2576 to and from BNs: it was completely broken. New compilation option
2577 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
2578 key elements as negative integers.
2581 *) Reorganize and speed up MD5.
2582 [Andy Polyakov <appro@fy.chalmers.se>]
2585 [Richard Levitte <richard@levitte.org>]
2587 *) New option -out to asn1parse to allow the parsed structure to be
2588 output to a file. This is most useful when combined with the -strparse
2589 option to examine the output of things like OCTET STRINGS.
2592 *) Make SSL library a little more fool-proof by not requiring any longer
2593 that SSL_set_{accept,connect}_state be called before
2594 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
2595 in many applications because usually everything *appeared* to work as
2596 intended anyway -- now it really works as intended).
2599 *) Move openssl.cnf out of lib/.
2602 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
2603 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
2604 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
2605 [Ralf S. Engelschall]
2607 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
2608 handle PKCS#7 enveloped data properly.
2609 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
2611 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
2612 copying pointers. The cert_st handling is changed by this in
2613 various ways (and thus what used to be known as ctx->default_cert
2614 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
2615 any longer when s->cert does not give us what we need).
2616 ssl_cert_instantiate becomes obsolete by this change.
2617 As soon as we've got the new code right (possibly it already is?),
2618 we have solved a couple of bugs of the earlier code where s->cert
2619 was used as if it could not have been shared with other SSL structures.
2621 Note that using the SSL API in certain dirty ways now will result
2622 in different behaviour than observed with earlier library versions:
2623 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
2624 does not influence s as it used to.
2626 In order to clean up things more thoroughly, inside SSL_SESSION
2627 we don't use CERT any longer, but a new structure SESS_CERT
2628 that holds per-session data (if available); currently, this is
2629 the peer's certificate chain and, for clients, the server's certificate
2630 and temporary key. CERT holds only those values that can have
2631 meaningful defaults in an SSL_CTX.
2634 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
2635 from the internal representation. Various PKCS#7 fixes: remove some
2636 evil casts and set the enc_dig_alg field properly based on the signing
2640 *) Allow PKCS#12 password to be set from the command line or the
2641 environment. Let 'ca' get its config file name from the environment
2642 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
2646 *) Allow certificate policies extension to use an IA5STRING for the
2647 organization field. This is contrary to the PKIX definition but
2648 VeriSign uses it and IE5 only recognises this form. Document 'x509'
2652 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
2653 without disallowing inline assembler and the like for non-pedantic builds.
2656 *) Support Borland C++ builder.
2657 [Janez Jere <jj@void.si>, modified by Ulf Möller]
2662 *) SHA-1 cleanups and performance enhancements.
2663 [Andy Polyakov <appro@fy.chalmers.se>]
2665 *) Sparc v8plus assembler for the bignum library.
2666 [Andy Polyakov <appro@fy.chalmers.se>]
2668 *) Accept any -xxx and +xxx compiler options in Configure.
2671 *) Update HPUX configuration.
2674 *) Add missing sk_<type>_unshift() function to safestack.h
2675 [Ralf S. Engelschall]
2677 *) New function SSL_CTX_use_certificate_chain_file that sets the
2678 "extra_cert"s in addition to the certificate. (This makes sense
2679 only for "PEM" format files, as chains as a whole are not
2683 *) Support verify_depth from the SSL API.
2684 x509_vfy.c had what can be considered an off-by-one-error:
2685 Its depth (which was not part of the external interface)
2686 was actually counting the number of certificates in a chain;
2687 now it really counts the depth.
2690 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
2691 instead of X509err, which often resulted in confusing error
2692 messages since the error codes are not globally unique
2693 (e.g. an alleged error in ssl3_accept when a certificate
2694 didn't match the private key).
2696 *) New function SSL_CTX_set_session_id_context that allows to set a default
2697 value (so that you don't need SSL_set_session_id_context for each
2698 connection using the SSL_CTX).
2701 *) OAEP decoding bug fix.
2704 *) Support INSTALL_PREFIX for package builders, as proposed by
2708 *) New Configure options "threads" and "no-threads". For systems
2709 where the proper compiler options are known (currently Solaris
2710 and Linux), "threads" is the default.
2713 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
2716 *) Install various scripts to $(OPENSSLDIR)/misc, not to
2717 $(INSTALLTOP)/bin -- they shouldn't clutter directories
2718 such as /usr/local/bin.
2721 *) "make linux-shared" to build shared libraries.
2722 [Niels Poppe <niels@netbox.org>]
2724 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
2727 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
2728 extension adding in x509 utility.
2731 *) Remove NOPROTO sections and error code comments.
2734 *) Partial rewrite of the DEF file generator to now parse the ANSI
2738 *) New Configure options --prefix=DIR and --openssldir=DIR.
2741 *) Complete rewrite of the error code script(s). It is all now handled
2742 by one script at the top level which handles error code gathering,
2743 header rewriting and C source file generation. It should be much better
2744 than the old method: it now uses a modified version of Ulf's parser to
2745 read the ANSI prototypes in all header files (thus the old K&R definitions
2746 aren't needed for error creation any more) and do a better job of
2747 translating function codes into names. The old 'ASN1 error code imbedded
2748 in a comment' is no longer necessary and it doesn't use .err files which
2749 have now been deleted. Also the error code call doesn't have to appear all
2750 on one line (which resulted in some large lines...).
2753 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
2756 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
2757 0 (which usually indicates a closed connection), but continue reading.
2760 *) Fix some race conditions.
2763 *) Add support for CRL distribution points extension. Add Certificate
2764 Policies and CRL distribution points documentation.
2767 *) Move the autogenerated header file parts to crypto/opensslconf.h.
2770 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
2771 8 of keying material. Merlin has also confirmed interop with this fix
2772 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
2773 [Merlin Hughes <merlin@baltimore.ie>]
2775 *) Fix lots of warnings.
2776 [Richard Levitte <levitte@stacken.kth.se>]
2778 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
2779 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
2780 [Richard Levitte <levitte@stacken.kth.se>]
2782 *) Fix problems with sizeof(long) == 8.
2783 [Andy Polyakov <appro@fy.chalmers.se>]
2785 *) Change functions to ANSI C.
2788 *) Fix typos in error codes.
2789 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
2791 *) Remove defunct assembler files from Configure.
2794 *) SPARC v8 assembler BIGNUM implementation.
2795 [Andy Polyakov <appro@fy.chalmers.se>]
2797 *) Support for Certificate Policies extension: both print and set.
2798 Various additions to support the r2i method this uses.
2801 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
2802 return a const string when you are expecting an allocated buffer.
2805 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
2806 types DirectoryString and DisplayText.
2809 *) Add code to allow r2i extensions to access the configuration database,
2810 add an LHASH database driver and add several ctx helper functions.
2813 *) Fix an evil bug in bn_expand2() which caused various BN functions to
2814 fail when they extended the size of a BIGNUM.
2817 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
2818 support typesafe stack.
2821 *) Fix typo in SSL_[gs]et_options().
2822 [Nils Frostberg <nils@medcom.se>]
2824 *) Delete various functions and files that belonged to the (now obsolete)
2825 old X509V3 handling code.
2828 *) New Configure option "rsaref".
2831 *) Don't auto-generate pem.h.
2834 *) Introduce type-safe ASN.1 SETs.
2837 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
2838 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
2840 *) Introduce type-safe STACKs. This will almost certainly break lots of code
2841 that links with OpenSSL (well at least cause lots of warnings), but fear
2842 not: the conversion is trivial, and it eliminates loads of evil casts. A
2843 few STACKed things have been converted already. Feel free to convert more.
2844 In the fullness of time, I'll do away with the STACK type altogether.
2847 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
2848 specified in <certfile> by updating the entry in the index.txt file.
2849 This way one no longer has to edit the index.txt file manually for
2850 revoking a certificate. The -revoke option does the gory details now.
2851 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
2853 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
2854 `-text' option at all and this way the `-noout -text' combination was
2855 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
2856 [Ralf S. Engelschall]
2858 *) Make sure a corresponding plain text error message exists for the
2859 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
2860 verify callback function determined that a certificate was revoked.
2861 [Ralf S. Engelschall]
2863 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
2864 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
2865 all available cipers including rc5, which was forgotten until now.
2866 In order to let the testing shell script know which algorithms
2867 are available, a new (up to now undocumented) command
2868 "openssl list-cipher-commands" is used.
2871 *) Bugfix: s_client occasionally would sleep in select() when
2872 it should have checked SSL_pending() first.
2875 *) New functions DSA_do_sign and DSA_do_verify to provide access to
2876 the raw DSA values prior to ASN.1 encoding.
2879 *) Tweaks to Configure
2880 [Niels Poppe <niels@netbox.org>]
2882 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
2886 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
2889 *) New config option to avoid instructions that are illegal on the 80386.
2890 The default code is faster, but requires at least a 486.
2893 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
2894 SSL2_SERVER_VERSION (not used at all) macros, which are now the
2895 same as SSL2_VERSION anyway.
2898 *) New "-showcerts" option for s_client.
2901 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2902 application. Various cleanups and fixes.
2905 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2906 modify error routines to work internally. Add error codes and PBE init
2907 to library startup routines.
2910 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2911 packing functions to asn1 and evp. Changed function names and error
2912 codes along the way.
2915 *) PKCS12 integration: and so it begins... First of several patches to
2916 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2917 objects to objects.h
2920 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2921 and display support for Thawte strong extranet extension.
2924 *) Add LinuxPPC support.
2925 [Jeff Dubrule <igor@pobox.org>]
2927 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2928 bn_div_words in alpha.s.
2929 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2931 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2932 OAEP isn't supported when OpenSSL is built with RSAref.
2933 [Ulf Moeller <ulf@fitug.de>]
2935 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2936 so they no longer are missing under -DNOPROTO.
2937 [Soren S. Jorvang <soren@t.dk>]
2940 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2942 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2943 doesn't work when the session is reused. Coming soon!
2946 *) Fix a security hole, that allows sessions to be reused in the wrong
2947 context thus bypassing client cert protection! All software that uses
2948 client certs and session caches in multiple contexts NEEDS PATCHING to
2949 allow session reuse! A fuller solution is in the works.
2950 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2952 *) Some more source tree cleanups (removed obsolete files
2953 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2954 permission on "config" script to be executable) and a fix for the INSTALL
2956 [Ulf Moeller <ulf@fitug.de>]
2958 *) Remove some legacy and erroneous uses of malloc, free instead of
2960 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2962 *) Make rsa_oaep_test return non-zero on error.
2963 [Ulf Moeller <ulf@fitug.de>]
2965 *) Add support for native Solaris shared libraries. Configure
2966 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2967 if someone would make that last step automatic.
2968 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2970 *) ctx_size was not built with the right compiler during "make links". Fixed.
2973 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2974 except NULL ciphers". This means the default cipher list will no longer
2975 enable NULL ciphers. They need to be specifically enabled e.g. with
2976 the string "DEFAULT:eNULL".
2979 *) Fix to RSA private encryption routines: if p < q then it would
2980 occasionally produce an invalid result. This will only happen with
2981 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2984 *) Be less restrictive and allow also `perl util/perlpath.pl
2985 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2986 because this way one can also use an interpreter named `perl5' (which is
2987 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2988 installed as `perl').
2989 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2991 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2992 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2994 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2995 advapi32.lib to Win32 build and change the pem test comparision
2996 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2997 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2998 and crypto/des/ede_cbcm_enc.c.
3001 *) DES quad checksum was broken on big-endian architectures. Fixed.
3004 *) Comment out two functions in bio.h that aren't implemented. Fix up the
3005 Win32 test batch file so it (might) work again. The Win32 test batch file
3006 is horrible: I feel ill....
3009 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
3010 in e_os.h. Audit of header files to check ANSI and non ANSI
3011 sections: 10 functions were absent from non ANSI section and not exported
3012 from Windows DLLs. Fixed up libeay.num for new functions.
3015 *) Make `openssl version' output lines consistent.
3016 [Ralf S. Engelschall]
3018 *) Fix Win32 symbol export lists for BIO functions: Added
3019 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
3020 to ms/libeay{16,32}.def.
3021 [Ralf S. Engelschall]
3023 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
3024 fine under Unix and passes some trivial tests I've now added. But the
3025 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
3026 added to make sure no one expects that this stuff really works in the
3027 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
3028 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
3030 [Ralf S. Engelschall]
3032 *) Fix the generation of two part addresses in perl.
3033 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
3035 *) Add config entry for Linux on MIPS.
3036 [John Tobey <jtobey@channel1.com>]
3038 *) Make links whenever Configure is run, unless we are on Windoze.
3041 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
3042 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
3046 *) Add a useful kludge to allow package maintainers to specify compiler and
3047 other platforms details on the command line without having to patch the
3048 Configure script everytime: One now can use ``perl Configure
3049 <id>:<details>'', i.e. platform ids are allowed to have details appended
3050 to them (seperated by colons). This is treated as there would be a static
3051 pre-configured entry in Configure's %table under key <id> with value
3052 <details> and ``perl Configure <id>'' is called. So, when you want to
3053 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
3054 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
3055 now, which overrides the FreeBSD-elf entry on-the-fly.
3056 [Ralf S. Engelschall]
3058 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
3061 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
3062 on the `perl Configure ...' command line. This way one can compile
3063 OpenSSL libraries with Position Independent Code (PIC) which is needed
3064 for linking it into DSOs.
3065 [Ralf S. Engelschall]
3067 *) Remarkably, export ciphers were totally broken and no-one had noticed!
3071 *) Cleaned up the LICENSE document: The official contact for any license
3072 questions now is the OpenSSL core team under openssl-core@openssl.org.
3073 And add a paragraph about the dual-license situation to make sure people
3074 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
3075 to the OpenSSL toolkit.
3076 [Ralf S. Engelschall]
3078 *) General source tree makefile cleanups: Made `making xxx in yyy...'
3079 display consistent in the source tree and replaced `/bin/rm' by `rm'.
3080 Additonally cleaned up the `make links' target: Remove unnecessary
3081 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
3082 to speed processing and no longer clutter the display with confusing
3083 stuff. Instead only the actually done links are displayed.
3084 [Ralf S. Engelschall]
3086 *) Permit null encryption ciphersuites, used for authentication only. It used
3087 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
3088 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
3092 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
3093 signed attributes when verifying signatures (this would break them),
3094 the detached data encoding was wrong and public keys obtained using
3095 X509_get_pubkey() weren't freed.
3098 *) Add text documentation for the BUFFER functions. Also added a work around
3099 to a Win95 console bug. This was triggered by the password read stuff: the
3100 last character typed gets carried over to the next fread(). If you were
3101 generating a new cert request using 'req' for example then the last
3102 character of the passphrase would be CR which would then enter the first
3106 *) Added the new `Includes OpenSSL Cryptography Software' button as
3107 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
3108 button and can be used by applications based on OpenSSL to show the
3109 relationship to the OpenSSL project.
3110 [Ralf S. Engelschall]
3112 *) Remove confusing variables in function signatures in files
3113 ssl/ssl_lib.c and ssl/ssl.h.
3114 [Lennart Bong <lob@kulthea.stacken.kth.se>]
3116 *) Don't install bss_file.c under PREFIX/include/
3117 [Lennart Bong <lob@kulthea.stacken.kth.se>]
3119 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
3120 functions that return function pointers and has support for NT specific
3121 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
3122 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
3123 unsigned to signed types: this was killing the Win32 compile.
3126 *) Add new certificate file to stack functions,
3127 SSL_add_dir_cert_subjects_to_stack() and
3128 SSL_add_file_cert_subjects_to_stack(). These largely supplant
3129 SSL_load_client_CA_file(), and can be used to add multiple certs easily
3130 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
3131 This means that Apache-SSL and similar packages don't have to mess around
3132 to add as many CAs as they want to the preferred list.
3135 *) Experiment with doxygen documentation. Currently only partially applied to
3137 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
3138 openssl.doxy as the configuration file.
3141 *) Get rid of remaining C++-style comments which strict C compilers hate.
3142 [Ralf S. Engelschall, pointed out by Carlos Amengual]
3144 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
3145 compiled in by default: it has problems with large keys.
3148 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
3149 DH private keys and/or callback functions which directly correspond to
3150 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
3151 is needed for applications which have to configure certificates on a
3152 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
3154 For the RSA certificate situation is makes no difference, but
3155 for the DSA certificate situation this fixes the "no shared cipher"
3156 problem where the OpenSSL cipher selection procedure failed because the
3157 temporary keys were not overtaken from the context and the API provided
3158 no way to reconfigure them.
3159 The new functions now let applications reconfigure the stuff and they
3160 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
3161 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
3162 non-public-API function ssl_cert_instantiate() is used as a helper
3163 function and also to reduce code redundancy inside ssl_rsa.c.
3164 [Ralf S. Engelschall]
3166 *) Move s_server -dcert and -dkey options out of the undocumented feature
3167 area because they are useful for the DSA situation and should be
3168 recognized by the users.
3169 [Ralf S. Engelschall]
3171 *) Fix the cipher decision scheme for export ciphers: the export bits are
3172 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
3173 SSL_EXP_MASK. So, the original variable has to be used instead of the
3174 already masked variable.
3175 [Richard Levitte <levitte@stacken.kth.se>]
3177 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
3178 [Richard Levitte <levitte@stacken.kth.se>]
3180 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
3181 from `int' to `unsigned int' because it's a length and initialized by
3182 EVP_DigestFinal() which expects an `unsigned int *'.
3183 [Richard Levitte <levitte@stacken.kth.se>]
3185 *) Don't hard-code path to Perl interpreter on shebang line of Configure
3186 script. Instead use the usual Shell->Perl transition trick.
3187 [Ralf S. Engelschall]
3189 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
3190 (in addition to RSA certificates) to match the behaviour of `openssl dsa
3191 -noout -modulus' as it's already the case for `openssl rsa -noout
3192 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
3193 currently the public key is printed (a decision which was already done by
3194 `openssl dsa -modulus' in the past) which serves a similar purpose.
3195 Additionally the NO_RSA no longer completely removes the whole -modulus
3196 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
3198 [Ralf S. Engelschall]
3200 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
3201 BIO. See the source (crypto/evp/bio_ok.c) for more info.
3202 [Arne Ansper <arne@ats.cyber.ee>]
3204 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
3205 to be added. Now both 'req' and 'ca' can use new objects defined in the
3209 *) Add cool BIO that does syslog (or event log on NT).
3210 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
3212 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
3213 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
3214 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
3215 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
3218 *) Add preliminary config info for new extension code.
3221 *) Make RSA_NO_PADDING really use no padding.
3222 [Ulf Moeller <ulf@fitug.de>]
3224 *) Generate errors when private/public key check is done.
3227 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
3228 for some CRL extensions and new objects added.
3231 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
3232 key usage extension and fuller support for authority key id.
3235 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
3236 padding method for RSA, which is recommended for new applications in PKCS
3237 #1 v2.0 (RFC 2437, October 1998).
3238 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
3239 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
3240 against Bleichbacher's attack on RSA.
3241 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
3244 *) Updates to the new SSL compression code
3245 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
3247 *) Fix so that the version number in the master secret, when passed
3248 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
3249 (because the server will not accept higher), that the version number
3250 is 0x03,0x01, not 0x03,0x00
3251 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
3253 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
3254 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
3255 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
3258 *) Support for RAW extensions where an arbitrary extension can be
3259 created by including its DER encoding. See apps/openssl.cnf for
3263 *) Make sure latest Perl versions don't interpret some generated C array
3264 code as Perl array code in the crypto/err/err_genc.pl script.
3265 [Lars Weber <3weber@informatik.uni-hamburg.de>]
3267 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
3268 not many people have the assembler. Various Win32 compilation fixes and
3269 update to the INSTALL.W32 file with (hopefully) more accurate Win32
3273 *) Modify configure script 'Configure' to automatically create crypto/date.h
3274 file under Win32 and also build pem.h from pem.org. New script
3275 util/mkfiles.pl to create the MINFO file on environments that can't do a
3276 'make files': perl util/mkfiles.pl >MINFO should work.
3279 *) Major rework of DES function declarations, in the pursuit of correctness
3280 and purity. As a result, many evil casts evaporated, and some weirdness,
3281 too. You may find this causes warnings in your code. Zapping your evil
3282 casts will probably fix them. Mostly.
3285 *) Fix for a typo in asn1.h. Bug fix to object creation script
3286 obj_dat.pl. It considered a zero in an object definition to mean
3287 "end of object": none of the objects in objects.h have any zeros
3288 so it wasn't spotted.
3289 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
3291 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
3292 Masking (CBCM). In the absence of test vectors, the best I have been able
3293 to do is check that the decrypt undoes the encrypt, so far. Send me test
3294 vectors if you have them.
3297 *) Correct calculation of key length for export ciphers (too much space was
3298 allocated for null ciphers). This has not been tested!
3301 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
3302 message is now correct (it understands "crypto" and "ssl" on its
3303 command line). There is also now an "update" option. This will update
3304 the util/ssleay.num and util/libeay.num files with any new functions.
3306 perl util/mkdef.pl crypto ssl update
3307 it will update them.
3310 *) Overhauled the Perl interface (perl/*):
3311 - ported BN stuff to OpenSSL's different BN library
3312 - made the perl/ source tree CVS-aware
3313 - renamed the package from SSLeay to OpenSSL (the files still contain
3314 their history because I've copied them in the repository)
3315 - removed obsolete files (the test scripts will be replaced
3316 by better Test::Harness variants in the future)
3317 [Ralf S. Engelschall]
3319 *) First cut for a very conservative source tree cleanup:
3320 1. merge various obsolete readme texts into doc/ssleay.txt
3321 where we collect the old documents and readme texts.
3322 2. remove the first part of files where I'm already sure that we no
3323 longer need them because of three reasons: either they are just temporary
3324 files which were left by Eric or they are preserved original files where
3325 I've verified that the diff is also available in the CVS via "cvs diff
3326 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
3327 the crypto/md/ stuff).
3328 [Ralf S. Engelschall]
3330 *) More extension code. Incomplete support for subject and issuer alt
3331 name, issuer and authority key id. Change the i2v function parameters
3332 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
3333 what that's for :-) Fix to ASN1 macro which messed up
3334 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
3337 *) Preliminary support for ENUMERATED type. This is largely copied from the
3341 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
3342 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
3344 *) Make sure `make rehash' target really finds the `openssl' program.
3345 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
3347 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
3348 like to hear about it if this slows down other processors.
3351 *) Add CygWin32 platform information to Configure script.
3352 [Alan Batie <batie@aahz.jf.intel.com>]
3354 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
3355 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
3357 *) New program nseq to manipulate netscape certificate sequences
3360 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
3364 *) Fixes to BN code. Previously the default was to define BN_RECURSION
3365 but the BN code had some problems that would cause failures when
3366 doing certificate verification and some other functions.
3367 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
3369 *) Add ASN1 and PEM code to support netscape certificate sequences.
3372 *) Add ASN1 and PEM code to support netscape certificate sequences.
3375 *) Add several PKIX and private extended key usage OIDs.
3378 *) Modify the 'ca' program to handle the new extension code. Modify
3379 openssl.cnf for new extension format, add comments.
3382 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
3383 and add a sample to openssl.cnf so req -x509 now adds appropriate
3387 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
3388 error code, add initial support to X509_print() and x509 application.
3391 *) Takes a deep breath and start addding X509 V3 extension support code. Add
3392 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
3393 stuff is currently isolated and isn't even compiled yet.
3396 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
3397 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
3398 Removed the versions check from X509 routines when loading extensions:
3399 this allows certain broken certificates that don't set the version
3400 properly to be processed.
3403 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
3404 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
3405 can still be regenerated with "make depend".
3408 *) Spelling mistake in C version of CAST-128.
3409 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
3411 *) Changes to the error generation code. The perl script err-code.pl
3412 now reads in the old error codes and retains the old numbers, only
3413 adding new ones if necessary. It also only changes the .err files if new
3414 codes are added. The makefiles have been modified to only insert errors
3415 when needed (to avoid needlessly modifying header files). This is done
3416 by only inserting errors if the .err file is newer than the auto generated
3417 C file. To rebuild all the error codes from scratch (the old behaviour)
3418 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
3419 or delete all the .err files.
3422 *) CAST-128 was incorrectly implemented for short keys. The C version has
3423 been fixed, but is untested. The assembler versions are also fixed, but
3424 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
3425 to regenerate it if needed.
3426 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
3427 Hagino <itojun@kame.net>]
3429 *) File was opened incorrectly in randfile.c.
3430 [Ulf Möller <ulf@fitug.de>]
3432 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
3433 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
3434 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
3435 al: it's just almost always a UTCTime. Note this patch adds new error
3436 codes so do a "make errors" if there are problems.
3439 *) Correct Linux 1 recognition in config.
3440 [Ulf Möller <ulf@fitug.de>]
3442 *) Remove pointless MD5 hash when using DSA keys in ca.
3443 [Anonymous <nobody@replay.com>]
3445 *) Generate an error if given an empty string as a cert directory. Also
3446 generate an error if handed NULL (previously returned 0 to indicate an
3447 error, but didn't set one).
3448 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
3450 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
3453 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
3454 parameters. This was causing a warning which killed off the Win32 compile.
3457 *) Remove C++ style comments from crypto/bn/bn_local.h.
3458 [Neil Costigan <neil.costigan@celocom.com>]
3460 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
3461 based on a text string, looking up short and long names and finally
3462 "dot" format. The "dot" format stuff didn't work. Added new function
3463 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
3464 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
3465 OID is not part of the table.
3468 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
3469 X509_LOOKUP_by_alias().
3472 *) Sort openssl functions by name.
3475 *) Get the gendsa program working (hopefully) and add it to app list. Remove
3476 encryption from sample DSA keys (in case anyone is interested the password
3480 *) Make _all_ *_free functions accept a NULL pointer.
3481 [Frans Heymans <fheymans@isaserver.be>]
3483 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
3485 [Anonymous <nobody@replay.com>]
3487 *) s_server should send the CAfile as acceptable CAs, not its own cert.
3488 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3490 *) Don't blow it for numeric -newkey arguments to apps/req.
3491 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3493 *) Temp key "for export" tests were wrong in s3_srvr.c.
3494 [Anonymous <nobody@replay.com>]
3496 *) Add prototype for temp key callback functions
3497 SSL_CTX_set_tmp_{rsa,dh}_callback().
3500 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
3501 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
3504 *) X509_name_add_entry() freed the wrong thing after an error.
3505 [Arne Ansper <arne@ats.cyber.ee>]
3507 *) rsa_eay.c would attempt to free a NULL context.
3508 [Arne Ansper <arne@ats.cyber.ee>]
3510 *) BIO_s_socket() had a broken should_retry() on Windoze.
3511 [Arne Ansper <arne@ats.cyber.ee>]
3513 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
3514 [Arne Ansper <arne@ats.cyber.ee>]
3516 *) Make sure the already existing X509_STORE->depth variable is initialized
3517 in X509_STORE_new(), but document the fact that this variable is still
3518 unused in the certificate verification process.
3519 [Ralf S. Engelschall]
3521 *) Fix the various library and apps files to free up pkeys obtained from
3522 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
3525 *) Fix reference counting in X509_PUBKEY_get(). This makes
3526 demos/maurice/example2.c work, amongst others, probably.
3527 [Steve Henson and Ben Laurie]
3529 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
3530 `openssl' and second, the shortcut symlinks for the `openssl <command>'
3531 are no longer created. This way we have a single and consistent command
3532 line interface `openssl <command>', similar to `cvs <command>'.
3533 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
3535 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
3536 BIT STRING wrapper always have zero unused bits.
3539 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
3542 *) Make the top-level INSTALL documentation easier to understand.
3545 *) Makefiles updated to exit if an error occurs in a sub-directory
3546 make (including if user presses ^C) [Paul Sutton]
3548 *) Make Montgomery context stuff explicit in RSA data structure.
3551 *) Fix build order of pem and err to allow for generated pem.h.
3554 *) Fix renumbering bug in X509_NAME_delete_entry().
3557 *) Enhanced the err-ins.pl script so it makes the error library number
3558 global and can add a library name. This is needed for external ASN1 and
3559 other error libraries.
3562 *) Fixed sk_insert which never worked properly.
3565 *) Fix ASN1 macros so they can handle indefinite length construted
3566 EXPLICIT tags. Some non standard certificates use these: they can now
3570 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
3571 into a single doc/ssleay.txt bundle. This way the information is still
3572 preserved but no longer messes up this directory. Now it's new room for
3573 the new set of documenation files.
3574 [Ralf S. Engelschall]
3576 *) SETs were incorrectly DER encoded. This was a major pain, because they
3577 shared code with SEQUENCEs, which aren't coded the same. This means that
3578 almost everything to do with SETs or SEQUENCEs has either changed name or
3579 number of arguments.
3580 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
3582 *) Fix test data to work with the above.
3585 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
3586 was already fixed by Eric for 0.9.1 it seems.
3587 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
3589 *) Autodetect FreeBSD3.
3592 *) Fix various bugs in Configure. This affects the following platforms:
3596 unixware-2.0-pentium
3600 *) Eliminate generated files from CVS. Reorder tests to regenerate files
3601 before they are needed.
3604 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
3608 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
3610 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
3611 changed SSLeay to OpenSSL in version strings.
3612 [Ralf S. Engelschall]
3614 *) Some fixups to the top-level documents.
3617 *) Fixed the nasty bug where rsaref.h was not found under compile-time
3618 because the symlink to include/ was missing.
3619 [Ralf S. Engelschall]
3621 *) Incorporated the popular no-RSA/DSA-only patches
3622 which allow to compile a RSA-free SSLeay.
3623 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
3625 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
3626 when "ssleay" is still not found.
3627 [Ralf S. Engelschall]
3629 *) Added more platforms to Configure: Cray T3E, HPUX 11,
3630 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
3632 *) Updated the README file.
3633 [Ralf S. Engelschall]
3635 *) Added various .cvsignore files in the CVS repository subdirs
3636 to make a "cvs update" really silent.
3637 [Ralf S. Engelschall]
3639 *) Recompiled the error-definition header files and added
3640 missing symbols to the Win32 linker tables.
3641 [Ralf S. Engelschall]
3643 *) Cleaned up the top-level documents;
3644 o new files: CHANGES and LICENSE
3645 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
3646 o merged COPYRIGHT into LICENSE
3647 o removed obsolete TODO file
3648 o renamed MICROSOFT to INSTALL.W32
3649 [Ralf S. Engelschall]
3651 *) Removed dummy files from the 0.9.1b source tree:
3652 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
3653 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
3654 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
3655 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
3656 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
3657 [Ralf S. Engelschall]
3659 *) Added various platform portability fixes.
3662 *) The Genesis of the OpenSSL rpject:
3663 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
3664 Young and Tim J. Hudson created while they were working for C2Net until
3666 [The OpenSSL Project]
3669 Changes between 0.9.0b and 0.9.1b [not released]
3671 *) Updated a few CA certificates under certs/
3674 *) Changed some BIGNUM api stuff.
3677 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
3678 DGUX x86, Linux Alpha, etc.
3681 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
3682 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
3686 *) Add -strparse option to asn1pars program which parses nested
3688 [Dr Stephen Henson <shenson@bigfoot.com>]
3690 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
3693 *) DSA fix for "ca" program.
3696 *) Added "-genkey" option to "dsaparam" program.
3699 *) Added RIPE MD160 (rmd160) message digest.
3702 *) Added -a (all) option to "ssleay version" command.
3705 *) Added PLATFORM define which is the id given to Configure.
3708 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
3711 *) Extended the ASN.1 parser routines.
3714 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
3717 *) Added a BN_CTX to the BN library.
3720 *) Fixed the weak key values in DES library
3723 *) Changed API in EVP library for cipher aliases.
3726 *) Added support for RC2/64bit cipher.
3729 *) Converted the lhash library to the crypto/mem.c functions.
3732 *) Added more recognized ASN.1 object ids.
3735 *) Added more RSA padding checks for SSL/TLS.
3738 *) Added BIO proxy/filter functionality.
3741 *) Added extra_certs to SSL_CTX which can be used
3742 send extra CA certificates to the client in the CA cert chain sending
3743 process. It can be configured with SSL_CTX_add_extra_chain_cert().
3746 *) Now Fortezza is denied in the authentication phase because
3747 this is key exchange mechanism is not supported by SSLeay at all.
3750 *) Additional PKCS1 checks.
3753 *) Support the string "TLSv1" for all TLS v1 ciphers.
3756 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
3757 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
3760 *) Fixed a few memory leaks.
3763 *) Fixed various code and comment typos.
3766 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
3767 bytes sent in the client random.
3768 [Edward Bishop <ebishop@spyglass.com>]