5 Changes between 0.9.5a and 0.9.6 [xx XXX 2000]
7 *) Add '-tls1' option to 'openssl ciphers', which was already
8 mentioned in the documentation but had not been implemented.
9 (This option is not yet really useful because even the additional
10 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
13 *) Initial DSO code added into libcrypto for letting OpenSSL (and
14 OpenSSL-based applications) load shared libraries and bind to
15 them in a portable way.
16 [Geoff Thorpe, with contributions from Richard Levitte]
18 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
20 *) Make sure _lrotl and _lrotr are only used with MSVC.
22 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
23 (the default implementation of RAND_status).
25 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
26 to '-clrext' (= clear extensions), as intended and documented.
27 [Bodo Moeller; inconsistency pointed out by Michael Attili
30 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
31 was larger than the MD block size.
32 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
34 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
35 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
36 using the passed key: if the passed key was a private key the result
37 of X509_print(), for example, would be to print out all the private key
41 *) des_quad_cksum() byte order bug fix.
42 [Ulf Möller, using the problem description in krb4-0.9.7, where
43 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
45 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
47 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
49 *) For easily testing in shell scripts whether some command
50 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
51 returns with exit code 0 iff no command of the given name is available.
52 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
53 the output goes to stdout and nothing is printed to stderr.
54 Additional arguments are always ignored.
56 Since for each cipher there is a command of the same name,
57 the 'no-cipher' compilation switches can be tested this way.
59 ('openssl no-XXX' is not able to detect pseudo-commands such
60 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
63 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
66 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
67 is set; it will be thrown away anyway because each handshake creates
69 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
70 to parameters -- in previous versions (since OpenSSL 0.9.3) the
71 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
72 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
75 *) New s_client option -ign_eof: EOF at stdin is ignored, and
76 'Q' and 'R' lose their special meanings (quit/renegotiate).
77 This is part of what -quiet does; unlike -quiet, -ign_eof
78 does not suppress any output.
81 *) Add compatibility options to the purpose and trust code. The
82 purpose X509_PURPOSE_ANY is "any purpose" which automatically
83 accepts a certificate or CA, this was the previous behaviour,
84 with all the associated security issues.
86 X509_TRUST_COMPAT is the old trust behaviour: only and
87 automatically trust self signed roots in certificate store. A
88 new trust setting X509_TRUST_DEFAULT is used to specify that
89 a purpose has no associated trust setting and it should instead
90 use the value in the default purpose.
93 *) Fix the PKCS#8 DSA private key code so it decodes keys again
94 and fix a memory leak.
97 *) In util/mkerr.pl (which implements 'make errors'), preserve
98 reason strings from the previous version of the .c file, as
99 the default to have only downcase letters (and digits) in
100 automatically generated reasons codes is not always appropriate.
103 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
104 using strerror. Previously, ERR_reason_error_string() returned
105 library names as reason strings for SYSerr; but SYSerr is a special
106 case where small numbers are errno values, not library numbers.
109 *) Add '-dsaparam' option to 'openssl dhparam' application. This
110 converts DSA parameters into DH parameters. (When creating parameters,
111 DSA_generate_parameters is used.)
114 *) Include 'length' (recommended exponent length) in C code generated
115 by 'openssl dhparam -C'.
118 *) The second argument to set_label in perlasm was already being used
119 so couldn't be used as a "file scope" flag. Moved to third argument
123 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
124 instead of RAND_bytes for encryption IVs and salts.
127 *) Include RAND_status() into RAND_METHOD instead of implementing
128 it only for md_rand.c Otherwise replacing the PRNG by calling
129 RAND_set_rand_method would be impossible.
132 *) Don't let DSA_generate_key() enter an infinite loop if the random
133 number generation fails.
136 *) New 'rand' application for creating pseudo-random output.
139 *) Added configuration support for Linux/IA64
140 [Rolf Haberrecker <rolf@suse.de>]
142 *) Assembler module support for Mingw32.
145 *) Shared library support for HPUX (in shlib/).
146 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
148 *) Shared library support for Solaris gcc.
149 [Lutz Behnke <behnke@trustcenter.de>]
151 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
153 *) PKCS7_encrypt() was adding text MIME headers twice because they
154 were added manually and by SMIME_crlf_copy().
157 *) In bntest.c don't call BN_rand with zero bits argument.
158 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
160 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
161 case was implemented. This caused BN_div_recp() to fail occasionally.
164 *) Add an optional second argument to the set_label() in the perl
165 assembly language builder. If this argument exists and is set
166 to 1 it signals that the assembler should use a symbol whose
167 scope is the entire file, not just the current function. This
168 is needed with MASM which uses the format label:: for this scope.
169 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
171 *) Change the ASN1 types so they are typedefs by default. Before
172 almost all types were #define'd to ASN1_STRING which was causing
173 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
177 *) Change names of new functions to the new get1/get0 naming
178 convention: After 'get1', the caller owns a reference count
179 and has to call ..._free; 'get0' returns a pointer to some
180 data structure without incrementing reference counters.
181 (Some of the existing 'get' functions increment a reference
182 counter, some don't.)
183 Similarly, 'set1' and 'add1' functions increase reference
184 counters or duplicate objects.
187 *) Allow for the possibility of temp RSA key generation failure:
188 the code used to assume it always worked and crashed on failure.
191 *) Fix potential buffer overrun problem in BIO_printf().
192 [Ulf Möller, using public domain code by Patrick Powell; problem
193 pointed out by David Sacerdote <das33@cornell.edu>]
195 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
196 RAND_egd() and RAND_status(). In the command line application,
197 the EGD socket can be specified like a seed file using RANDFILE
201 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
202 Some CAs (e.g. Verisign) distribute certificates in this form.
205 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
206 list to exclude them. This means that no special compilation option
207 is needed to use anonymous DH: it just needs to be included in the
211 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
212 EVP_MD_type. The old functionality is available in a new macro called
213 EVP_MD_md(). Change code that uses it and update docs.
216 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
217 where the 'void *' argument is replaced by a function pointer argument.
218 Previously 'void *' was abused to point to functions, which works on
219 many platforms, but is not correct. As these functions are usually
220 called by macros defined in OpenSSL header files, most source code
221 should work without changes.
224 *) <openssl/opensslconf.h> (which is created by Configure) now contains
225 sections with information on -D... compiler switches used for
226 compiling the library so that applications can see them. To enable
227 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
228 must be defined. E.g.,
229 #define OPENSSL_ALGORITHM_DEFINES
230 #include <openssl/opensslconf.h>
231 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
232 [Richard Levitte, Ulf and Bodo Möller]
234 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
238 *) Change the 'other' type in certificate aux info to a STACK_OF
239 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
240 the required ASN1 format: arbitrary types determined by an OID.
243 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
244 argument to 'req'. This is not because the function is newer or
245 better than others it just uses the work 'NEW' in the certificate
246 request header lines. Some software needs this.
249 *) Reorganise password command line arguments: now passwords can be
250 obtained from various sources. Delete the PEM_cb function and make
251 it the default behaviour: i.e. if the callback is NULL and the
252 usrdata argument is not NULL interpret it as a null terminated pass
253 phrase. If usrdata and the callback are NULL then the pass phrase
254 is prompted for as usual.
257 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
258 the support is automatically enabled. The resulting binaries will
259 autodetect the card and use it if present.
260 [Ben Laurie and Compaq Inc.]
262 *) Work around for Netscape hang bug. This sends certificate request
263 and server done in one record. Since this is perfectly legal in the
264 SSL/TLS protocol it isn't a "bug" option and is on by default. See
265 the bugs/SSLv3 entry for more info.
268 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
271 *) Add -rand argument to smime and pkcs12 applications and read/write
275 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
278 *) Add command line password options to the remaining applications.
281 *) Bug fix for BN_div_recp() for numerators with an even number of
285 *) More tests in bntest.c, and changed test_bn output.
288 *) ./config recognizes MacOS X now.
291 *) Bug fix for BN_div() when the first words of num and divsor are
292 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
295 *) Add support for various broken PKCS#8 formats, and command line
296 options to produce them.
299 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
300 get temporary BIGNUMs from a BN_CTX.
303 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
307 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
308 include a #define from the old name to the new. The original intent
309 was that statically linked binaries could for example just call
310 SSLeay_add_all_ciphers() to just add ciphers to the table and not
311 link with digests. This never worked becayse SSLeay_add_all_digests()
312 and SSLeay_add_all_ciphers() were in the same source file so calling
313 one would link with the other. They are now in separate source files.
316 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
319 *) Use a less unusual form of the Miller-Rabin primality test (it used
320 a binary algorithm for exponentiation integrated into the Miller-Rabin
321 loop, our standard modexp algorithms are faster).
324 *) Support for the EBCDIC character set completed.
325 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
327 *) Source code cleanups: use const where appropriate, eliminate casts,
328 use void * instead of char * in lhash.
331 *) Bugfix: ssl3_send_server_key_exchange was not restartable
332 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
333 this the server could overwrite ephemeral keys that the client
337 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
338 using 50 iterations of the Rabin-Miller test.
340 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
341 iterations of the Rabin-Miller test as required by the appendix
342 to FIPS PUB 186[-1]) instead of DSA_is_prime.
343 As BN_is_prime_fasttest includes trial division, DSA parameter
344 generation becomes much faster.
346 This implies a change for the callback functions in DSA_is_prime
347 and DSA_generate_parameters: The callback function is called once
348 for each positive witness in the Rabin-Miller test, not just
349 occasionally in the inner loop; and the parameters to the
350 callback function now provide an iteration count for the outer
351 loop rather than for the current invocation of the inner loop.
352 DSA_generate_parameters additionally can call the callback
353 function with an 'iteration count' of -1, meaning that a
354 candidate has passed the trial division test (when q is generated
355 from an application-provided seed, trial division is skipped).
358 *) New function BN_is_prime_fasttest that optionally does trial
359 division before starting the Rabin-Miller test and has
360 an additional BN_CTX * argument (whereas BN_is_prime always
361 has to allocate at least one BN_CTX).
362 'callback(1, -1, cb_arg)' is called when a number has passed the
363 trial division stage.
366 *) Fix for bug in CRL encoding. The validity dates weren't being handled
370 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
373 *) New function BN_pseudo_rand().
376 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
377 bignum version of BN_from_montgomery() with the working code from
378 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
382 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
383 made it impossible to use the same SSL_SESSION data structure in
384 SSL2 clients in multiple threads.
387 *) The return value of RAND_load_file() no longer counts bytes obtained
388 by stat(). RAND_load_file(..., -1) is new and uses the complete file
389 to seed the PRNG (previously an explicit byte count was required).
390 [Ulf Möller, Bodo Möller]
392 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
393 used (char *) instead of (void *) and had casts all over the place.
396 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
399 *) Retain source code compatibility for BN_prime_checks macro:
400 BN_is_prime(..., BN_prime_checks, ...) now uses
401 BN_prime_checks_for_size to determine the appropriate number of
402 Rabin-Miller iterations.
405 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
406 DH_CHECK_P_NOT_SAFE_PRIME.
407 (Check if this is true? OpenPGP calls them "strong".)
410 *) Merge the functionality of "dh" and "gendh" programs into a new program
411 "dhparam". The old programs are retained for now but will handle DH keys
412 (instead of parameters) in future.
415 *) Make the ciphers, s_server and s_client programs check the return values
416 when a new cipher list is set.
419 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
420 ciphers. Before when the 56bit ciphers were enabled the sorting was
423 The syntax for the cipher sorting has been extended to support sorting by
424 cipher-strength (using the strength_bits hard coded in the tables).
425 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
427 Fix a bug in the cipher-command parser: when supplying a cipher command
428 string with an "undefined" symbol (neither command nor alphanumeric
429 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
432 Due to the strength-sorting extension, the code of the
433 ssl_create_cipher_list() function was completely rearranged. I hope that
434 the readability was also increased :-)
435 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
437 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
438 for the first serial number and places 2 in the serial number file. This
439 avoids problems when the root CA is created with serial number zero and
440 the first user certificate has the same issuer name and serial number
444 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
445 the new code. Add documentation for this stuff.
448 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
449 X509_*() to X509at_*() on the grounds that they don't handle X509
450 structures and behave in an analagous way to the X509v3 functions:
451 they shouldn't be called directly but wrapper functions should be used
454 So we also now have some wrapper functions that call the X509at functions
455 when passed certificate requests. (TO DO: similar things can be done with
456 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
457 things. Some of these need some d2i or i2d and print functionality
458 because they handle more complex structures.)
461 *) Add missing #ifndefs that caused missing symbols when building libssl
462 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
464 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
466 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
467 has a return value which indicates the quality of the random data
468 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
469 error queue. New function RAND_pseudo_bytes() generates output that is
470 guaranteed to be unique but not unpredictable. RAND_add is like
471 RAND_seed, but takes an extra argument for an entropy estimate
472 (RAND_seed always assumes full entropy).
475 *) Do more iterations of Rabin-Miller probable prime test (specifically,
476 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
477 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
478 in crypto/bn/bn_prime.c for the complete table). This guarantees a
479 false-positive rate of at most 2^-80 for random input.
482 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
485 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
486 in the 0.9.5 release), this returns the chain
487 from an X509_CTX structure with a dup of the stack and all
488 the X509 reference counts upped: so the stack will exist
489 after X509_CTX_cleanup() has been called. Modify pkcs12.c
492 Also make SSL_SESSION_print() print out the verify return
496 *) Add manpage for the pkcs12 command. Also change the default
497 behaviour so MAC iteration counts are used unless the new
498 -nomaciter option is used. This improves file security and
499 only older versions of MSIE (4.0 for example) need it.
502 *) Honor the no-xxx Configure options when creating .DEF files.
505 *) Add PKCS#10 attributes to field table: challengePassword,
506 unstructuredName and unstructuredAddress. These are taken from
507 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
508 international characters are used.
510 More changes to X509_ATTRIBUTE code: allow the setting of types
511 based on strings. Remove the 'loc' parameter when adding
512 attributes because these will be a SET OF encoding which is sorted
516 *) Initial changes to the 'req' utility to allow request generation
517 automation. This will allow an application to just generate a template
518 file containing all the field values and have req construct the
521 Initial support for X509_ATTRIBUTE handling. Stacks of these are
522 used all over the place including certificate requests and PKCS#7
523 structures. They are currently handled manually where necessary with
524 some primitive wrappers for PKCS#7. The new functions behave in a
525 manner analogous to the X509 extension functions: they allow
526 attributes to be looked up by NID and added.
528 Later something similar to the X509V3 code would be desirable to
529 automatically handle the encoding, decoding and printing of the
530 more complex types. The string types like challengePassword can
531 be handled by the string table functions.
533 Also modified the multi byte string table handling. Now there is
534 a 'global mask' which masks out certain types. The table itself
535 can use the flag STABLE_NO_MASK to ignore the mask setting: this
536 is useful when for example there is only one permissible type
537 (as in countryName) and using the mask might result in no valid
541 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
542 SSL_get_peer_finished to allow applications to obtain the latest
543 Finished messages sent to the peer or expected from the peer,
544 respectively. (SSL_get_peer_finished is usually the Finished message
545 actually received from the peer, otherwise the protocol will be aborted.)
547 As the Finished message are message digests of the complete handshake
548 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
549 be used for external authentication procedures when the authentication
550 provided by SSL/TLS is not desired or is not enough.
553 *) Enhanced support for Alpha Linux is added. Now ./config checks if
554 the host supports BWX extension and if Compaq C is present on the
555 $PATH. Just exploiting of the BWX extension results in 20-30%
556 performance kick for some algorithms, e.g. DES and RC4 to mention
557 a couple. Compaq C in turn generates ~20% faster code for MD5 and
561 *) Add support for MS "fast SGC". This is arguably a violation of the
562 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
563 weak crypto and after checking the certificate is SGC a second one
564 with strong crypto. MS SGC stops the first handshake after receiving
565 the server certificate message and sends a second client hello. Since
566 a server will typically do all the time consuming operations before
567 expecting any further messages from the client (server key exchange
568 is the most expensive) there is little difference between the two.
570 To get OpenSSL to support MS SGC we have to permit a second client
571 hello message after we have sent server done. In addition we have to
572 reset the MAC if we do get this second client hello.
575 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
576 if a DER encoded private key is RSA or DSA traditional format. Changed
577 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
578 format DER encoded private key. Newer code should use PKCS#8 format which
579 has the key type encoded in the ASN1 structure. Added DER private key
580 support to pkcs8 application.
583 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
584 ciphersuites has been selected (as required by the SSL 3/TLS 1
585 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
586 is set, we interpret this as a request to violate the specification
587 (the worst that can happen is a handshake failure, and 'correct'
588 behaviour would result in a handshake failure anyway).
591 *) In SSL_CTX_add_session, take into account that there might be multiple
592 SSL_SESSION structures with the same session ID (e.g. when two threads
593 concurrently obtain them from an external cache).
594 The internal cache can handle only one SSL_SESSION with a given ID,
595 so if there's a conflict, we now throw out the old one to achieve
599 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
600 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
601 some routines that use cipher OIDs: some ciphers do not have OIDs
602 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
606 *) Simplify the trust setting structure and code. Now we just have
607 two sequences of OIDs for trusted and rejected settings. These will
608 typically have values the same as the extended key usage extension
609 and any application specific purposes.
611 The trust checking code now has a default behaviour: it will just
612 check for an object with the same NID as the passed id. Functions can
613 be provided to override either the default behaviour or the behaviour
614 for a given id. SSL client, server and email already have functions
615 in place for compatibility: they check the NID and also return "trusted"
616 if the certificate is self signed.
619 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
620 traditional format into an EVP_PKEY structure.
623 *) Add a password callback function PEM_cb() which either prompts for
624 a password if usr_data is NULL or otherwise assumes it is a null
625 terminated password. Allow passwords to be passed on command line
626 environment or config files in a few more utilities.
629 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
630 keys. Add some short names for PKCS#8 PBE algorithms and allow them
631 to be specified on the command line for the pkcs8 and pkcs12 utilities.
632 Update documentation.
635 *) Support for ASN1 "NULL" type. This could be handled before by using
636 ASN1_TYPE but there wasn't any function that would try to read a NULL
637 and produce an error if it couldn't. For compatibility we also have
638 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
639 don't allocate anything because they don't need to.
642 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
644 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
646 *) Rebuild of the memory allocation routines used by OpenSSL code and
647 possibly others as well. The purpose is to make an interface that
648 provide hooks so anyone can build a separate set of allocation and
649 deallocation routines to be used by OpenSSL, for example memory
650 pool implementations, or something else, which was previously hard
651 since Malloc(), Realloc() and Free() were defined as macros having
652 the values malloc, realloc and free, respectively (except for Win32
653 compilations). The same is provided for memory debugging code.
654 OpenSSL already comes with functionality to find memory leaks, but
655 this gives people a chance to debug other memory problems.
657 With these changes, a new set of functions and macros have appeared:
659 CRYPTO_set_mem_debug_functions() [F]
660 CRYPTO_get_mem_debug_functions() [F]
661 CRYPTO_dbg_set_options() [F]
662 CRYPTO_dbg_get_options() [F]
663 CRYPTO_malloc_debug_init() [M]
665 The memory debug functions are NULL by default, unless the library
666 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
667 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
668 gives the standard debugging functions that come with OpenSSL) or
669 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
670 provided by the library user) must be used. When the standard
671 debugging functions are used, CRYPTO_dbg_set_options can be used to
672 request additional information:
673 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
674 the CRYPTO_MDEBUG_xxx macro when compiling the library.
676 Also, things like CRYPTO_set_mem_functions will always give the
677 expected result (the new set of functions is used for allocation
678 and deallocation) at all times, regardless of platform and compiler
681 To finish it up, some functions that were never use in any other
682 way than through macros have a new API and new semantic:
688 All macros of value have retained their old syntax.
689 [Richard Levitte and Bodo Moeller]
691 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
692 ordering of SMIMECapabilities wasn't in "strength order" and there
693 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
697 *) Some ASN1 types with illegal zero length encoding (INTEGER,
698 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
699 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
701 *) Merge in my S/MIME library for OpenSSL. This provides a simple
702 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
703 functionality to handle multipart/signed properly) and a utility
704 called 'smime' to call all this stuff. This is based on code I
705 originally wrote for Celo who have kindly allowed it to be
709 *) Add variants des_set_key_checked and des_set_key_unchecked of
710 des_set_key (aka des_key_sched). Global variable des_check_key
711 decides which of these is called by des_set_key; this way
712 des_check_key behaves as it always did, but applications and
713 the library itself, which was buggy for des_check_key == 1,
714 have a cleaner way to pick the version they need.
717 *) New function PKCS12_newpass() which changes the password of a
721 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
722 dynamic mix. In both cases the ids can be used as an index into the
723 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
724 functions so they accept a list of the field values and the
725 application doesn't need to directly manipulate the X509_TRUST
729 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
733 *) Modify the way the V3 extension code looks up extensions. This now
734 works in a similar way to the object code: we have some "standard"
735 extensions in a static table which is searched with OBJ_bsearch()
736 and the application can add dynamic ones if needed. The file
737 crypto/x509v3/ext_dat.h now has the info: this file needs to be
738 updated whenever a new extension is added to the core code and kept
739 in ext_nid order. There is a simple program 'tabtest.c' which checks
740 this. New extensions are not added too often so this file can readily
741 be maintained manually.
743 There are two big advantages in doing things this way. The extensions
744 can be looked up immediately and no longer need to be "added" using
745 X509V3_add_standard_extensions(): this function now does nothing.
746 [Side note: I get *lots* of email saying the extension code doesn't
747 work because people forget to call this function]
748 Also no dynamic allocation is done unless new extensions are added:
749 so if we don't add custom extensions there is no need to call
750 X509V3_EXT_cleanup().
753 *) Modify enc utility's salting as follows: make salting the default. Add a
754 magic header, so unsalted files fail gracefully instead of just decrypting
755 to garbage. This is because not salting is a big security hole, so people
756 should be discouraged from doing it.
759 *) Fixes and enhancements to the 'x509' utility. It allowed a message
760 digest to be passed on the command line but it only used this
761 parameter when signing a certificate. Modified so all relevant
762 operations are affected by the digest parameter including the
763 -fingerprint and -x509toreq options. Also -x509toreq choked if a
764 DSA key was used because it didn't fix the digest.
767 *) Initial certificate chain verify code. Currently tests the untrusted
768 certificates for consistency with the verify purpose (which is set
769 when the X509_STORE_CTX structure is set up) and checks the pathlength.
771 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
772 this is because it will reject chains with invalid extensions whereas
773 every previous version of OpenSSL and SSLeay made no checks at all.
775 Trust code: checks the root CA for the relevant trust settings. Trust
776 settings have an initial value consistent with the verify purpose: e.g.
777 if the verify purpose is for SSL client use it expects the CA to be
778 trusted for SSL client use. However the default value can be changed to
779 permit custom trust settings: one example of this would be to only trust
780 certificates from a specific "secure" set of CAs.
782 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
783 which should be used for version portability: especially since the
784 verify structure is likely to change more often now.
786 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
787 to set them. If not set then assume SSL clients will verify SSL servers
790 Two new options to the verify program: -untrusted allows a set of
791 untrusted certificates to be passed in and -purpose which sets the
792 intended purpose of the certificate. If a purpose is set then the
793 new chain verify code is used to check extension consistency.
796 *) Support for the authority information access extension.
799 *) Modify RSA and DSA PEM read routines to transparently handle
800 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
801 public keys in a format compatible with certificate
802 SubjectPublicKeyInfo structures. Unfortunately there were already
803 functions called *_PublicKey_* which used various odd formats so
804 these are retained for compatibility: however the DSA variants were
805 never in a public release so they have been deleted. Changed dsa/rsa
806 utilities to handle the new format: note no releases ever handled public
807 keys so we should be OK.
809 The primary motivation for this change is to avoid the same fiasco
810 that dogs private keys: there are several incompatible private key
811 formats some of which are standard and some OpenSSL specific and
812 require various evil hacks to allow partial transparent handling and
813 even then it doesn't work with DER formats. Given the option anything
814 other than PKCS#8 should be dumped: but the other formats have to
815 stay in the name of compatibility.
817 With public keys and the benefit of hindsight one standard format
818 is used which works with EVP_PKEY, RSA or DSA structures: though
819 it clearly returns an error if you try to read the wrong kind of key.
821 Added a -pubkey option to the 'x509' utility to output the public key.
822 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
823 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
824 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
825 that do the same as the EVP_PKEY_assign_*() except they up the
826 reference count of the added key (they don't "swallow" the
830 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
831 CRLs would fail if the file contained no certificates or no CRLs:
832 added a new function to read in both types and return the number
833 read: this means that if none are read it will be an error. The
834 DER versions of the certificate and CRL reader would always fail
835 because it isn't possible to mix certificates and CRLs in DER format
836 without choking one or the other routine. Changed this to just read
837 a certificate: this is the best we can do. Also modified the code
838 in apps/verify.c to take notice of return codes: it was previously
839 attempting to read in certificates from NULL pointers and ignoring
840 any errors: this is one reason why the cert and CRL reader seemed
841 to work. It doesn't check return codes from the default certificate
842 routines: these may well fail if the certificates aren't installed.
845 *) Code to support otherName option in GeneralName.
848 *) First update to verify code. Change the verify utility
849 so it warns if it is passed a self signed certificate:
850 for consistency with the normal behaviour. X509_verify
851 has been modified to it will now verify a self signed
852 certificate if *exactly* the same certificate appears
853 in the store: it was previously impossible to trust a
854 single self signed certificate. This means that:
855 openssl verify ss.pem
856 now gives a warning about a self signed certificate but
857 openssl verify -CAfile ss.pem ss.pem
861 *) For servers, store verify_result in SSL_SESSION data structure
862 (and add it to external session representation).
863 This is needed when client certificate verifications fails,
864 but an application-provided verification callback (set by
865 SSL_CTX_set_cert_verify_callback) allows accepting the session
866 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
867 but returns 1): When the session is reused, we have to set
868 ssl->verify_result to the appropriate error code to avoid
870 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
872 *) Fix a bug in the new PKCS#7 code: it didn't consider the
873 case in PKCS7_dataInit() where the signed PKCS7 structure
874 didn't contain any existing data because it was being created.
875 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
877 *) Add a salt to the key derivation routines in enc.c. This
878 forms the first 8 bytes of the encrypted file. Also add a
879 -S option to allow a salt to be input on the command line.
882 *) New function X509_cmp(). Oddly enough there wasn't a function
883 to compare two certificates. We do this by working out the SHA1
884 hash and comparing that. X509_cmp() will be needed by the trust
888 *) SSL_get1_session() is like SSL_get_session(), but increments
889 the reference count in the SSL_SESSION returned.
890 [Geoff Thorpe <geoff@eu.c2.net>]
892 *) Fix for 'req': it was adding a null to request attributes.
893 Also change the X509_LOOKUP and X509_INFO code to handle
894 certificate auxiliary information.
897 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
901 *) Add the possibility to add extra information to the memory leak
902 detecting output, to form tracebacks, showing from where each
903 allocation was originated: CRYPTO_push_info("constant string") adds
904 the string plus current file name and line number to a per-thread
905 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
906 is like calling CYRPTO_pop_info() until the stack is empty.
907 Also updated memory leak detection code to be multi-thread-safe.
910 *) Add options -text and -noout to pkcs7 utility and delete the
911 encryption options which never did anything. Update docs.
914 *) Add options to some of the utilities to allow the pass phrase
915 to be included on either the command line (not recommended on
916 OSes like Unix) or read from the environment. Update the
917 manpages and fix a few bugs.
920 *) Add a few manpages for some of the openssl commands.
923 *) Fix the -revoke option in ca. It was freeing up memory twice,
924 leaking and not finding already revoked certificates.
927 *) Extensive changes to support certificate auxiliary information.
928 This involves the use of X509_CERT_AUX structure and X509_AUX
929 functions. An X509_AUX function such as PEM_read_X509_AUX()
930 can still read in a certificate file in the usual way but it
931 will also read in any additional "auxiliary information". By
932 doing things this way a fair degree of compatibility can be
933 retained: existing certificates can have this information added
934 using the new 'x509' options.
936 Current auxiliary information includes an "alias" and some trust
937 settings. The trust settings will ultimately be used in enhanced
938 certificate chain verification routines: currently a certificate
939 can only be trusted if it is self signed and then it is trusted
943 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
944 The problem was that one of the replacement routines had not been working
945 since SSLeay releases. For now the offending routine has been replaced
946 with non-optimised assembler. Even so, this now gives around 95%
947 performance improvement for 1024 bit RSA signs.
950 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
951 handling. Most clients have the effective key size in bits equal to
952 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
953 A few however don't do this and instead use the size of the decrypted key
954 to determine the RC2 key length and the AlgorithmIdentifier to determine
955 the effective key length. In this case the effective key length can still
956 be 40 bits but the key length can be 168 bits for example. This is fixed
957 by manually forcing an RC2 key into the EVP_PKEY structure because the
958 EVP code can't currently handle unusual RC2 key sizes: it always assumes
959 the key length and effective key length are equal.
962 *) Add a bunch of functions that should simplify the creation of
963 X509_NAME structures. Now you should be able to do:
964 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
965 and have it automatically work out the correct field type and fill in
966 the structures. The more adventurous can try:
967 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
968 and it will (hopefully) work out the correct multibyte encoding.
971 *) Change the 'req' utility to use the new field handling and multibyte
972 copy routines. Before the DN field creation was handled in an ad hoc
973 way in req, ca, and x509 which was rather broken and didn't support
974 BMPStrings or UTF8Strings. Since some software doesn't implement
975 BMPStrings or UTF8Strings yet, they can be enabled using the config file
976 using the dirstring_type option. See the new comment in the default
977 openssl.cnf for more info.
980 *) Make crypto/rand/md_rand.c more robust:
981 - Assure unique random numbers after fork().
982 - Make sure that concurrent threads access the global counter and
983 md serializably so that we never lose entropy in them
984 or use exactly the same state in multiple threads.
985 Access to the large state is not always serializable because
986 the additional locking could be a performance killer, and
987 md should be large enough anyway.
990 *) New file apps/app_rand.c with commonly needed functionality
991 for handling the random seed file.
993 Use the random seed file in some applications that previously did not:
995 dsaparam -genkey (which also ignored its '-rand' option),
999 Except on systems with /dev/urandom, it is crucial to have a random
1000 seed file at least for key creation, DSA signing, and for DH exchanges;
1001 for RSA signatures we could do without one.
1003 gendh and gendsa (unlike genrsa) used to read only the first byte
1004 of each file listed in the '-rand' option. The function as previously
1005 found in genrsa is now in app_rand.c and is used by all programs
1006 that support '-rand'.
1009 *) In RAND_write_file, use mode 0600 for creating files;
1010 don't just chmod when it may be too late.
1013 *) Report an error from X509_STORE_load_locations
1014 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1017 *) New function ASN1_mbstring_copy() this copies a string in either
1018 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1019 into an ASN1_STRING type. A mask of permissible types is passed
1020 and it chooses the "minimal" type to use or an error if not type
1024 *) Add function equivalents to the various macros in asn1.h. The old
1025 macros are retained with an M_ prefix. Code inside the library can
1026 use the M_ macros. External code (including the openssl utility)
1027 should *NOT* in order to be "shared library friendly".
1030 *) Add various functions that can check a certificate's extensions
1031 to see if it usable for various purposes such as SSL client,
1032 server or S/MIME and CAs of these types. This is currently
1033 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1034 verification. Also added a -purpose flag to x509 utility to
1035 print out all the purposes.
1038 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1042 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1043 for, obtain and decode and extension and obtain its critical flag.
1044 This allows all the necessary extension code to be handled in a
1045 single function call.
1048 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1049 platforms. See crypto/rc4/rc4_enc.c for further details.
1052 *) New -noout option to asn1parse. This causes no output to be produced
1053 its main use is when combined with -strparse and -out to extract data
1054 from a file (which may not be in ASN.1 format).
1057 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1058 when producing the local key id.
1059 [Richard Levitte <levitte@stacken.kth.se>]
1061 *) New option -dhparam in s_server. This allows a DH parameter file to be
1062 stated explicitly. If it is not stated then it tries the first server
1063 certificate file. The previous behaviour hard coded the filename
1067 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1068 a public key to be input or output. For example:
1069 openssl rsa -in key.pem -pubout -out pubkey.pem
1070 Also added necessary DSA public key functions to handle this.
1073 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1074 in the message. This was handled by allowing
1075 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1076 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1078 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1079 to the end of the strings whereas this didn't. This would cause problems
1080 if strings read with d2i_ASN1_bytes() were later modified.
1081 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1083 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1084 data and it contains EOF it will end up returning an error. This is
1085 caused by input 46 bytes long. The cause is due to the way base64
1086 BIOs find the start of base64 encoded data. They do this by trying a
1087 trial decode on each line until they find one that works. When they
1088 do a flag is set and it starts again knowing it can pass all the
1089 data directly through the decoder. Unfortunately it doesn't reset
1090 the context it uses. This means that if EOF is reached an attempt
1091 is made to pass two EOFs through the context and this causes the
1092 resulting error. This can also cause other problems as well. As is
1093 usual with these problems it takes *ages* to find and the fix is
1094 trivial: move one line.
1095 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1097 *) Ugly workaround to get s_client and s_server working under Windows. The
1098 old code wouldn't work because it needed to select() on sockets and the
1099 tty (for keypresses and to see if data could be written). Win32 only
1100 supports select() on sockets so we select() with a 1s timeout on the
1101 sockets and then see if any characters are waiting to be read, if none
1102 are present then we retry, we also assume we can always write data to
1103 the tty. This isn't nice because the code then blocks until we've
1104 received a complete line of data and it is effectively polling the
1105 keyboard at 1s intervals: however it's quite a bit better than not
1106 working at all :-) A dedicated Windows application might handle this
1107 with an event loop for example.
1110 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
1111 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
1112 will be called when RSA_sign() and RSA_verify() are used. This is useful
1113 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
1114 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
1115 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
1116 This necessitated the support of an extra signature type NID_md5_sha1
1117 for SSL signatures and modifications to the SSL library to use it instead
1118 of calling RSA_public_decrypt() and RSA_private_encrypt().
1121 *) Add new -verify -CAfile and -CApath options to the crl program, these
1122 will lookup a CRL issuers certificate and verify the signature in a
1123 similar way to the verify program. Tidy up the crl program so it
1124 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
1125 less strict. It will now permit CRL extensions even if it is not
1126 a V2 CRL: this will allow it to tolerate some broken CRLs.
1129 *) Initialize all non-automatic variables each time one of the openssl
1130 sub-programs is started (this is necessary as they may be started
1131 multiple times from the "OpenSSL>" prompt).
1132 [Lennart Bang, Bodo Moeller]
1134 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
1135 removing all other RSA functionality (this is what NO_RSA does). This
1136 is so (for example) those in the US can disable those operations covered
1137 by the RSA patent while allowing storage and parsing of RSA keys and RSA
1141 *) Non-copying interface to BIO pairs.
1142 (still largely untested)
1145 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
1146 ASCII string. This was handled independently in various places before.
1149 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
1150 UTF8 strings a character at a time.
1153 *) Use client_version from client hello to select the protocol
1154 (s23_srvr.c) and for RSA client key exchange verification
1155 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
1158 *) Add various utility functions to handle SPKACs, these were previously
1159 handled by poking round in the structure internals. Added new function
1160 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
1161 print, verify and generate SPKACs. Based on an original idea from
1162 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
1165 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
1168 *) Allow the config file extension section to be overwritten on the
1169 command line. Based on an original idea from Massimiliano Pala
1170 <madwolf@comune.modena.it>. The new option is called -extensions
1171 and can be applied to ca, req and x509. Also -reqexts to override
1172 the request extensions in req and -crlexts to override the crl extensions
1176 *) Add new feature to the SPKAC handling in ca. Now you can include
1177 the same field multiple times by preceding it by "XXXX." for example:
1180 this is the same syntax as used in the req config file.
1183 *) Allow certificate extensions to be added to certificate requests. These
1184 are specified in a 'req_extensions' option of the req section of the
1185 config file. They can be printed out with the -text option to req but
1186 are otherwise ignored at present.
1189 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
1190 data read consists of only the final block it would not decrypted because
1191 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
1192 A misplaced 'break' also meant the decrypted final block might not be
1193 copied until the next read.
1196 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
1197 a few extra parameters to the DH structure: these will be useful if
1198 for example we want the value of 'q' or implement X9.42 DH.
1201 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
1202 provides hooks that allow the default DSA functions or functions on a
1203 "per key" basis to be replaced. This allows hardware acceleration and
1204 hardware key storage to be handled without major modification to the
1205 library. Also added low level modexp hooks and CRYPTO_EX structure and
1206 associated functions.
1209 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
1210 as "read only": it can't be written to and the buffer it points to will
1211 not be freed. Reading from a read only BIO is much more efficient than
1212 a normal memory BIO. This was added because there are several times when
1213 an area of memory needs to be read from a BIO. The previous method was
1214 to create a memory BIO and write the data to it, this results in two
1215 copies of the data and an O(n^2) reading algorithm. There is a new
1216 function BIO_new_mem_buf() which creates a read only memory BIO from
1217 an area of memory. Also modified the PKCS#7 routines to use read only
1221 *) Bugfix: ssl23_get_client_hello did not work properly when called in
1222 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
1223 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
1224 but a retry condition occured while trying to read the rest.
1227 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
1228 NID_pkcs7_encrypted by default: this was wrong since this should almost
1229 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
1230 the encrypted data type: this is a more sensible place to put it and it
1231 allows the PKCS#12 code to be tidied up that duplicated this
1235 *) Changed obj_dat.pl script so it takes its input and output files on
1236 the command line. This should avoid shell escape redirection problems
1240 *) Initial support for certificate extension requests, these are included
1241 in things like Xenroll certificate requests. Included functions to allow
1242 extensions to be obtained and added.
1245 *) -crlf option to s_client and s_server for sending newlines as
1246 CRLF (as required by many protocols).
1249 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
1251 *) Install libRSAglue.a when OpenSSL is built with RSAref.
1252 [Ralf S. Engelschall]
1254 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
1255 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
1257 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
1261 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
1262 DH parameters/keys (q is lost during that conversion, but the resulting
1263 DH parameters contain its length).
1265 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
1266 much faster than DH_generate_parameters (which creates parameters
1267 where p = 2*q + 1), and also the smaller q makes DH computations
1268 much more efficient (160-bit exponentiation instead of 1024-bit
1269 exponentiation); so this provides a convenient way to support DHE
1270 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
1271 utter importance to use
1272 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1274 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1275 when such DH parameters are used, because otherwise small subgroup
1276 attacks may become possible!
1279 *) Avoid memory leak in i2d_DHparams.
1282 *) Allow the -k option to be used more than once in the enc program:
1283 this allows the same encrypted message to be read by multiple recipients.
1286 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
1287 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
1288 it will always use the numerical form of the OID, even if it has a short
1292 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
1293 method only got called if p,q,dmp1,dmq1,iqmp components were present,
1294 otherwise bn_mod_exp was called. In the case of hardware keys for example
1295 no private key components need be present and it might store extra data
1296 in the RSA structure, which cannot be accessed from bn_mod_exp.
1297 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
1298 private key operations.
1301 *) Added support for SPARC Linux.
1304 *) pem_password_cb function type incompatibly changed from
1305 typedef int pem_password_cb(char *buf, int size, int rwflag);
1307 ....(char *buf, int size, int rwflag, void *userdata);
1308 so that applications can pass data to their callbacks:
1309 The PEM[_ASN1]_{read,write}... functions and macros now take an
1310 additional void * argument, which is just handed through whenever
1311 the password callback is called.
1312 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
1314 New function SSL_CTX_set_default_passwd_cb_userdata.
1316 Compatibility note: As many C implementations push function arguments
1317 onto the stack in reverse order, the new library version is likely to
1318 interoperate with programs that have been compiled with the old
1319 pem_password_cb definition (PEM_whatever takes some data that
1320 happens to be on the stack as its last argument, and the callback
1321 just ignores this garbage); but there is no guarantee whatsoever that
1324 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
1325 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
1326 problems not only on Windows, but also on some Unix platforms.
1327 To avoid problematic command lines, these definitions are now in an
1328 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
1329 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
1332 *) MIPS III/IV assembler module is reimplemented.
1335 *) More DES library cleanups: remove references to srand/rand and
1336 delete an unused file.
1339 *) Add support for the the free Netwide assembler (NASM) under Win32,
1340 since not many people have MASM (ml) and it can be hard to obtain.
1341 This is currently experimental but it seems to work OK and pass all
1342 the tests. Check out INSTALL.W32 for info.
1345 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
1346 without temporary keys kept an extra copy of the server key,
1347 and connections with temporary keys did not free everything in case
1351 *) New function RSA_check_key and new openssl rsa option -check
1352 for verifying the consistency of RSA keys.
1353 [Ulf Moeller, Bodo Moeller]
1355 *) Various changes to make Win32 compile work:
1356 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
1357 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
1358 comparison" warnings.
1359 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
1362 *) Add a debugging option to PKCS#5 v2 key generation function: when
1363 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
1364 derived keys are printed to stderr.
1367 *) Copy the flags in ASN1_STRING_dup().
1368 [Roman E. Pavlov <pre@mo.msk.ru>]
1370 *) The x509 application mishandled signing requests containing DSA
1371 keys when the signing key was also DSA and the parameters didn't match.
1373 It was supposed to omit the parameters when they matched the signing key:
1374 the verifying software was then supposed to automatically use the CA's
1375 parameters if they were absent from the end user certificate.
1377 Omitting parameters is no longer recommended. The test was also
1378 the wrong way round! This was probably due to unusual behaviour in
1379 EVP_cmp_parameters() which returns 1 if the parameters match.
1380 This meant that parameters were omitted when they *didn't* match and
1381 the certificate was useless. Certificates signed with 'ca' didn't have
1383 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1385 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1386 The interface is as follows:
1387 Applications can use
1388 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1389 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1390 "off" is now the default.
1391 The library internally uses
1392 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1393 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1394 to disable memory-checking temporarily.
1396 Some inconsistent states that previously were possible (and were
1397 even the default) are now avoided.
1399 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1400 with each memory chunk allocated; this is occasionally more helpful
1401 than just having a counter.
1403 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1405 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1409 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1410 which largely parallels "options", but is for changing API behaviour,
1411 whereas "options" are about protocol behaviour.
1412 Initial "mode" flags are:
1414 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1415 a single record has been written.
1416 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1417 retries use the same buffer location.
1418 (But all of the contents must be
1422 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1425 *) Fix problems with no-hmac etc.
1426 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1428 *) New functions RSA_get_default_method(), RSA_set_method() and
1429 RSA_get_method(). These allows replacement of RSA_METHODs without having
1430 to mess around with the internals of an RSA structure.
1433 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1434 Also really enable memory leak checks in openssl.c and in some
1436 [Chad C. Mulligan, Bodo Moeller]
1438 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1439 up the length of negative integers. This has now been simplified to just
1440 store the length when it is first determined and use it later, rather
1441 than trying to keep track of where data is copied and updating it to
1443 [Steve Henson, reported by Brien Wheeler
1444 <bwheeler@authentica-security.com>]
1446 *) Add a new function PKCS7_signatureVerify. This allows the verification
1447 of a PKCS#7 signature but with the signing certificate passed to the
1448 function itself. This contrasts with PKCS7_dataVerify which assumes the
1449 certificate is present in the PKCS#7 structure. This isn't always the
1450 case: certificates can be omitted from a PKCS#7 structure and be
1451 distributed by "out of band" means (such as a certificate database).
1454 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1455 function prototypes in pem.h, also change util/mkdef.pl to add the
1456 necessary function names.
1459 *) mk1mf.pl (used by Windows builds) did not properly read the
1460 options set by Configure in the top level Makefile, and Configure
1461 was not even able to write more than one option correctly.
1462 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1465 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1466 file to be loaded from a BIO or FILE pointer. The BIO version will
1467 for example allow memory BIOs to contain config info.
1470 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1471 Whoever hopes to achieve shared-library compatibility across versions
1472 must use this, not the compile-time macro.
1473 (Exercise 0.9.4: Which is the minimum library version required by
1475 Note: All this applies only to multi-threaded programs, others don't
1479 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1480 through a BIO pair triggered the default case, i.e.
1481 SSLerr(...,SSL_R_UNKNOWN_STATE).
1484 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1485 can use the SSL library even if none of the specific BIOs is
1489 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1490 for the encoded length.
1491 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1493 *) Add initial documentation of the X509V3 functions.
1496 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1497 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1498 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1499 secure PKCS#8 private key format with a high iteration count.
1502 *) Fix determination of Perl interpreter: A perl or perl5
1503 _directory_ in $PATH was also accepted as the interpreter.
1504 [Ralf S. Engelschall]
1506 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1507 wrong with it but it was very old and did things like calling
1508 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1512 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1513 to use the new extension code.
1516 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1517 with macros. This should make it easier to change their form, add extra
1518 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1522 *) Add to configuration table a new entry that can specify an alternative
1523 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1524 according to Mark Crispin <MRC@Panda.COM>.
1528 *) DES CBC did not update the IV. Weird.
1531 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1532 Changing the behaviour of the former might break existing programs --
1533 where IV updating is needed, des_ncbc_encrypt can be used.
1536 *) When bntest is run from "make test" it drives bc to check its
1537 calculations, as well as internally checking them. If an internal check
1538 fails, it needs to cause bc to give a non-zero result or make test carries
1539 on without noticing the failure. Fixed.
1542 *) DES library cleanups.
1545 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
1546 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
1547 ciphers. NOTE: although the key derivation function has been verified
1548 against some published test vectors it has not been extensively tested
1549 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
1553 *) Instead of "mkdir -p", which is not fully portable, use new
1554 Perl script "util/mkdir-p.pl".
1557 *) Rewrite the way password based encryption (PBE) is handled. It used to
1558 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
1559 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
1560 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
1561 the 'parameter' field of the AlgorithmIdentifier is passed to the
1562 underlying key generation function so it must do its own ASN1 parsing.
1563 This has also changed the EVP_PBE_CipherInit() function which now has a
1564 'parameter' argument instead of literal salt and iteration count values
1565 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
1568 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
1569 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
1570 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
1571 KEY" because this clashed with PKCS#8 unencrypted string. Since this
1572 value was just used as a "magic string" and not used directly its
1573 value doesn't matter.
1576 *) Introduce some semblance of const correctness to BN. Shame C doesn't
1580 *) "linux-sparc64" configuration (ultrapenguin).
1581 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
1582 "linux-sparc" configuration.
1583 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
1585 *) config now generates no-xxx options for missing ciphers.
1588 *) Support the EBCDIC character set (work in progress).
1589 File ebcdic.c not yet included because it has a different license.
1590 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1592 *) Support BS2000/OSD-POSIX.
1593 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1595 *) Make callbacks for key generation use void * instead of char *.
1598 *) Make S/MIME samples compile (not yet tested).
1601 *) Additional typesafe stacks.
1604 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
1608 Changes between 0.9.3 and 0.9.3a [29 May 1999]
1610 *) New configuration variant "sco5-gcc".
1612 *) Updated some demos.
1613 [Sean O Riordain, Wade Scholine]
1615 *) Add missing BIO_free at exit of pkcs12 application.
1618 *) Fix memory leak in conf.c.
1621 *) Updates for Win32 to assembler version of MD5.
1624 *) Set #! path to perl in apps/der_chop to where we found it
1625 instead of using a fixed path.
1628 *) SHA library changes for irix64-mips4-cc.
1631 *) Improvements for VMS support.
1635 Changes between 0.9.2b and 0.9.3 [24 May 1999]
1637 *) Bignum library bug fix. IRIX 6 passes "make test" now!
1638 This also avoids the problems with SC4.2 and unpatched SC5.
1639 [Andy Polyakov <appro@fy.chalmers.se>]
1641 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
1642 These are required because of the typesafe stack would otherwise break
1643 existing code. If old code used a structure member which used to be STACK
1644 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
1645 sk_num or sk_value it would produce an error because the num, data members
1646 are not present in STACK_OF. Now it just produces a warning. sk_set
1647 replaces the old method of assigning a value to sk_value
1648 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
1649 that does this will no longer work (and should use sk_set instead) but
1650 this could be regarded as a "questionable" behaviour anyway.
1653 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
1654 correctly handle encrypted S/MIME data.
1657 *) Change type of various DES function arguments from des_cblock
1658 (which means, in function argument declarations, pointer to char)
1659 to des_cblock * (meaning pointer to array with 8 char elements),
1660 which allows the compiler to do more typechecking; it was like
1661 that back in SSLeay, but with lots of ugly casts.
1663 Introduce new type const_des_cblock.
1666 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
1667 problems: find RecipientInfo structure that matches recipient certificate
1668 and initialise the ASN1 structures properly based on passed cipher.
1671 *) Belatedly make the BN tests actually check the results.
1674 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
1675 to and from BNs: it was completely broken. New compilation option
1676 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
1677 key elements as negative integers.
1680 *) Reorganize and speed up MD5.
1681 [Andy Polyakov <appro@fy.chalmers.se>]
1684 [Richard Levitte <richard@levitte.org>]
1686 *) New option -out to asn1parse to allow the parsed structure to be
1687 output to a file. This is most useful when combined with the -strparse
1688 option to examine the output of things like OCTET STRINGS.
1691 *) Make SSL library a little more fool-proof by not requiring any longer
1692 that SSL_set_{accept,connect}_state be called before
1693 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
1694 in many applications because usually everything *appeared* to work as
1695 intended anyway -- now it really works as intended).
1698 *) Move openssl.cnf out of lib/.
1701 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
1702 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
1703 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
1704 [Ralf S. Engelschall]
1706 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
1707 handle PKCS#7 enveloped data properly.
1708 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
1710 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
1711 copying pointers. The cert_st handling is changed by this in
1712 various ways (and thus what used to be known as ctx->default_cert
1713 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
1714 any longer when s->cert does not give us what we need).
1715 ssl_cert_instantiate becomes obsolete by this change.
1716 As soon as we've got the new code right (possibly it already is?),
1717 we have solved a couple of bugs of the earlier code where s->cert
1718 was used as if it could not have been shared with other SSL structures.
1720 Note that using the SSL API in certain dirty ways now will result
1721 in different behaviour than observed with earlier library versions:
1722 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
1723 does not influence s as it used to.
1725 In order to clean up things more thoroughly, inside SSL_SESSION
1726 we don't use CERT any longer, but a new structure SESS_CERT
1727 that holds per-session data (if available); currently, this is
1728 the peer's certificate chain and, for clients, the server's certificate
1729 and temporary key. CERT holds only those values that can have
1730 meaningful defaults in an SSL_CTX.
1733 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
1734 from the internal representation. Various PKCS#7 fixes: remove some
1735 evil casts and set the enc_dig_alg field properly based on the signing
1739 *) Allow PKCS#12 password to be set from the command line or the
1740 environment. Let 'ca' get its config file name from the environment
1741 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
1745 *) Allow certificate policies extension to use an IA5STRING for the
1746 organization field. This is contrary to the PKIX definition but
1747 VeriSign uses it and IE5 only recognises this form. Document 'x509'
1751 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
1752 without disallowing inline assembler and the like for non-pedantic builds.
1755 *) Support Borland C++ builder.
1756 [Janez Jere <jj@void.si>, modified by Ulf Möller]
1761 *) SHA-1 cleanups and performance enhancements.
1762 [Andy Polyakov <appro@fy.chalmers.se>]
1764 *) Sparc v8plus assembler for the bignum library.
1765 [Andy Polyakov <appro@fy.chalmers.se>]
1767 *) Accept any -xxx and +xxx compiler options in Configure.
1770 *) Update HPUX configuration.
1773 *) Add missing sk_<type>_unshift() function to safestack.h
1774 [Ralf S. Engelschall]
1776 *) New function SSL_CTX_use_certificate_chain_file that sets the
1777 "extra_cert"s in addition to the certificate. (This makes sense
1778 only for "PEM" format files, as chains as a whole are not
1782 *) Support verify_depth from the SSL API.
1783 x509_vfy.c had what can be considered an off-by-one-error:
1784 Its depth (which was not part of the external interface)
1785 was actually counting the number of certificates in a chain;
1786 now it really counts the depth.
1789 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
1790 instead of X509err, which often resulted in confusing error
1791 messages since the error codes are not globally unique
1792 (e.g. an alleged error in ssl3_accept when a certificate
1793 didn't match the private key).
1795 *) New function SSL_CTX_set_session_id_context that allows to set a default
1796 value (so that you don't need SSL_set_session_id_context for each
1797 connection using the SSL_CTX).
1800 *) OAEP decoding bug fix.
1803 *) Support INSTALL_PREFIX for package builders, as proposed by
1807 *) New Configure options "threads" and "no-threads". For systems
1808 where the proper compiler options are known (currently Solaris
1809 and Linux), "threads" is the default.
1812 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
1815 *) Install various scripts to $(OPENSSLDIR)/misc, not to
1816 $(INSTALLTOP)/bin -- they shouldn't clutter directories
1817 such as /usr/local/bin.
1820 *) "make linux-shared" to build shared libraries.
1821 [Niels Poppe <niels@netbox.org>]
1823 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
1826 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
1827 extension adding in x509 utility.
1830 *) Remove NOPROTO sections and error code comments.
1833 *) Partial rewrite of the DEF file generator to now parse the ANSI
1837 *) New Configure options --prefix=DIR and --openssldir=DIR.
1840 *) Complete rewrite of the error code script(s). It is all now handled
1841 by one script at the top level which handles error code gathering,
1842 header rewriting and C source file generation. It should be much better
1843 than the old method: it now uses a modified version of Ulf's parser to
1844 read the ANSI prototypes in all header files (thus the old K&R definitions
1845 aren't needed for error creation any more) and do a better job of
1846 translating function codes into names. The old 'ASN1 error code imbedded
1847 in a comment' is no longer necessary and it doesn't use .err files which
1848 have now been deleted. Also the error code call doesn't have to appear all
1849 on one line (which resulted in some large lines...).
1852 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
1855 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
1856 0 (which usually indicates a closed connection), but continue reading.
1859 *) Fix some race conditions.
1862 *) Add support for CRL distribution points extension. Add Certificate
1863 Policies and CRL distribution points documentation.
1866 *) Move the autogenerated header file parts to crypto/opensslconf.h.
1869 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
1870 8 of keying material. Merlin has also confirmed interop with this fix
1871 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
1872 [Merlin Hughes <merlin@baltimore.ie>]
1874 *) Fix lots of warnings.
1875 [Richard Levitte <levitte@stacken.kth.se>]
1877 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
1878 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
1879 [Richard Levitte <levitte@stacken.kth.se>]
1881 *) Fix problems with sizeof(long) == 8.
1882 [Andy Polyakov <appro@fy.chalmers.se>]
1884 *) Change functions to ANSI C.
1887 *) Fix typos in error codes.
1888 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
1890 *) Remove defunct assembler files from Configure.
1893 *) SPARC v8 assembler BIGNUM implementation.
1894 [Andy Polyakov <appro@fy.chalmers.se>]
1896 *) Support for Certificate Policies extension: both print and set.
1897 Various additions to support the r2i method this uses.
1900 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
1901 return a const string when you are expecting an allocated buffer.
1904 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
1905 types DirectoryString and DisplayText.
1908 *) Add code to allow r2i extensions to access the configuration database,
1909 add an LHASH database driver and add several ctx helper functions.
1912 *) Fix an evil bug in bn_expand2() which caused various BN functions to
1913 fail when they extended the size of a BIGNUM.
1916 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
1917 support typesafe stack.
1920 *) Fix typo in SSL_[gs]et_options().
1921 [Nils Frostberg <nils@medcom.se>]
1923 *) Delete various functions and files that belonged to the (now obsolete)
1924 old X509V3 handling code.
1927 *) New Configure option "rsaref".
1930 *) Don't auto-generate pem.h.
1933 *) Introduce type-safe ASN.1 SETs.
1936 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
1937 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
1939 *) Introduce type-safe STACKs. This will almost certainly break lots of code
1940 that links with OpenSSL (well at least cause lots of warnings), but fear
1941 not: the conversion is trivial, and it eliminates loads of evil casts. A
1942 few STACKed things have been converted already. Feel free to convert more.
1943 In the fullness of time, I'll do away with the STACK type altogether.
1946 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
1947 specified in <certfile> by updating the entry in the index.txt file.
1948 This way one no longer has to edit the index.txt file manually for
1949 revoking a certificate. The -revoke option does the gory details now.
1950 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
1952 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
1953 `-text' option at all and this way the `-noout -text' combination was
1954 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
1955 [Ralf S. Engelschall]
1957 *) Make sure a corresponding plain text error message exists for the
1958 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
1959 verify callback function determined that a certificate was revoked.
1960 [Ralf S. Engelschall]
1962 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
1963 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
1964 all available cipers including rc5, which was forgotten until now.
1965 In order to let the testing shell script know which algorithms
1966 are available, a new (up to now undocumented) command
1967 "openssl list-cipher-commands" is used.
1970 *) Bugfix: s_client occasionally would sleep in select() when
1971 it should have checked SSL_pending() first.
1974 *) New functions DSA_do_sign and DSA_do_verify to provide access to
1975 the raw DSA values prior to ASN.1 encoding.
1978 *) Tweaks to Configure
1979 [Niels Poppe <niels@netbox.org>]
1981 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
1985 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
1988 *) New config option to avoid instructions that are illegal on the 80386.
1989 The default code is faster, but requires at least a 486.
1992 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
1993 SSL2_SERVER_VERSION (not used at all) macros, which are now the
1994 same as SSL2_VERSION anyway.
1997 *) New "-showcerts" option for s_client.
2000 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2001 application. Various cleanups and fixes.
2004 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2005 modify error routines to work internally. Add error codes and PBE init
2006 to library startup routines.
2009 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2010 packing functions to asn1 and evp. Changed function names and error
2011 codes along the way.
2014 *) PKCS12 integration: and so it begins... First of several patches to
2015 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2016 objects to objects.h
2019 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2020 and display support for Thawte strong extranet extension.
2023 *) Add LinuxPPC support.
2024 [Jeff Dubrule <igor@pobox.org>]
2026 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2027 bn_div_words in alpha.s.
2028 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2030 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2031 OAEP isn't supported when OpenSSL is built with RSAref.
2032 [Ulf Moeller <ulf@fitug.de>]
2034 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2035 so they no longer are missing under -DNOPROTO.
2036 [Soren S. Jorvang <soren@t.dk>]
2039 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2041 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2042 doesn't work when the session is reused. Coming soon!
2045 *) Fix a security hole, that allows sessions to be reused in the wrong
2046 context thus bypassing client cert protection! All software that uses
2047 client certs and session caches in multiple contexts NEEDS PATCHING to
2048 allow session reuse! A fuller solution is in the works.
2049 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2051 *) Some more source tree cleanups (removed obsolete files
2052 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2053 permission on "config" script to be executable) and a fix for the INSTALL
2055 [Ulf Moeller <ulf@fitug.de>]
2057 *) Remove some legacy and erroneous uses of malloc, free instead of
2059 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2061 *) Make rsa_oaep_test return non-zero on error.
2062 [Ulf Moeller <ulf@fitug.de>]
2064 *) Add support for native Solaris shared libraries. Configure
2065 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2066 if someone would make that last step automatic.
2067 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2069 *) ctx_size was not built with the right compiler during "make links". Fixed.
2072 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2073 except NULL ciphers". This means the default cipher list will no longer
2074 enable NULL ciphers. They need to be specifically enabled e.g. with
2075 the string "DEFAULT:eNULL".
2078 *) Fix to RSA private encryption routines: if p < q then it would
2079 occasionally produce an invalid result. This will only happen with
2080 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2083 *) Be less restrictive and allow also `perl util/perlpath.pl
2084 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2085 because this way one can also use an interpreter named `perl5' (which is
2086 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2087 installed as `perl').
2088 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2090 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2091 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2093 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2094 advapi32.lib to Win32 build and change the pem test comparision
2095 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2096 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2097 and crypto/des/ede_cbcm_enc.c.
2100 *) DES quad checksum was broken on big-endian architectures. Fixed.
2103 *) Comment out two functions in bio.h that aren't implemented. Fix up the
2104 Win32 test batch file so it (might) work again. The Win32 test batch file
2105 is horrible: I feel ill....
2108 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
2109 in e_os.h. Audit of header files to check ANSI and non ANSI
2110 sections: 10 functions were absent from non ANSI section and not exported
2111 from Windows DLLs. Fixed up libeay.num for new functions.
2114 *) Make `openssl version' output lines consistent.
2115 [Ralf S. Engelschall]
2117 *) Fix Win32 symbol export lists for BIO functions: Added
2118 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
2119 to ms/libeay{16,32}.def.
2120 [Ralf S. Engelschall]
2122 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
2123 fine under Unix and passes some trivial tests I've now added. But the
2124 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
2125 added to make sure no one expects that this stuff really works in the
2126 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
2127 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
2129 [Ralf S. Engelschall]
2131 *) Fix the generation of two part addresses in perl.
2132 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
2134 *) Add config entry for Linux on MIPS.
2135 [John Tobey <jtobey@channel1.com>]
2137 *) Make links whenever Configure is run, unless we are on Windoze.
2140 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
2141 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
2145 *) Add a useful kludge to allow package maintainers to specify compiler and
2146 other platforms details on the command line without having to patch the
2147 Configure script everytime: One now can use ``perl Configure
2148 <id>:<details>'', i.e. platform ids are allowed to have details appended
2149 to them (seperated by colons). This is treated as there would be a static
2150 pre-configured entry in Configure's %table under key <id> with value
2151 <details> and ``perl Configure <id>'' is called. So, when you want to
2152 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
2153 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
2154 now, which overrides the FreeBSD-elf entry on-the-fly.
2155 [Ralf S. Engelschall]
2157 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
2160 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
2161 on the `perl Configure ...' command line. This way one can compile
2162 OpenSSL libraries with Position Independent Code (PIC) which is needed
2163 for linking it into DSOs.
2164 [Ralf S. Engelschall]
2166 *) Remarkably, export ciphers were totally broken and no-one had noticed!
2170 *) Cleaned up the LICENSE document: The official contact for any license
2171 questions now is the OpenSSL core team under openssl-core@openssl.org.
2172 And add a paragraph about the dual-license situation to make sure people
2173 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
2174 to the OpenSSL toolkit.
2175 [Ralf S. Engelschall]
2177 *) General source tree makefile cleanups: Made `making xxx in yyy...'
2178 display consistent in the source tree and replaced `/bin/rm' by `rm'.
2179 Additonally cleaned up the `make links' target: Remove unnecessary
2180 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
2181 to speed processing and no longer clutter the display with confusing
2182 stuff. Instead only the actually done links are displayed.
2183 [Ralf S. Engelschall]
2185 *) Permit null encryption ciphersuites, used for authentication only. It used
2186 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
2187 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
2191 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
2192 signed attributes when verifying signatures (this would break them),
2193 the detached data encoding was wrong and public keys obtained using
2194 X509_get_pubkey() weren't freed.
2197 *) Add text documentation for the BUFFER functions. Also added a work around
2198 to a Win95 console bug. This was triggered by the password read stuff: the
2199 last character typed gets carried over to the next fread(). If you were
2200 generating a new cert request using 'req' for example then the last
2201 character of the passphrase would be CR which would then enter the first
2205 *) Added the new `Includes OpenSSL Cryptography Software' button as
2206 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
2207 button and can be used by applications based on OpenSSL to show the
2208 relationship to the OpenSSL project.
2209 [Ralf S. Engelschall]
2211 *) Remove confusing variables in function signatures in files
2212 ssl/ssl_lib.c and ssl/ssl.h.
2213 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2215 *) Don't install bss_file.c under PREFIX/include/
2216 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2218 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
2219 functions that return function pointers and has support for NT specific
2220 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
2221 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
2222 unsigned to signed types: this was killing the Win32 compile.
2225 *) Add new certificate file to stack functions,
2226 SSL_add_dir_cert_subjects_to_stack() and
2227 SSL_add_file_cert_subjects_to_stack(). These largely supplant
2228 SSL_load_client_CA_file(), and can be used to add multiple certs easily
2229 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
2230 This means that Apache-SSL and similar packages don't have to mess around
2231 to add as many CAs as they want to the preferred list.
2234 *) Experiment with doxygen documentation. Currently only partially applied to
2236 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
2237 openssl.doxy as the configuration file.
2240 *) Get rid of remaining C++-style comments which strict C compilers hate.
2241 [Ralf S. Engelschall, pointed out by Carlos Amengual]
2243 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
2244 compiled in by default: it has problems with large keys.
2247 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
2248 DH private keys and/or callback functions which directly correspond to
2249 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
2250 is needed for applications which have to configure certificates on a
2251 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
2253 For the RSA certificate situation is makes no difference, but
2254 for the DSA certificate situation this fixes the "no shared cipher"
2255 problem where the OpenSSL cipher selection procedure failed because the
2256 temporary keys were not overtaken from the context and the API provided
2257 no way to reconfigure them.
2258 The new functions now let applications reconfigure the stuff and they
2259 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
2260 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
2261 non-public-API function ssl_cert_instantiate() is used as a helper
2262 function and also to reduce code redundancy inside ssl_rsa.c.
2263 [Ralf S. Engelschall]
2265 *) Move s_server -dcert and -dkey options out of the undocumented feature
2266 area because they are useful for the DSA situation and should be
2267 recognized by the users.
2268 [Ralf S. Engelschall]
2270 *) Fix the cipher decision scheme for export ciphers: the export bits are
2271 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
2272 SSL_EXP_MASK. So, the original variable has to be used instead of the
2273 already masked variable.
2274 [Richard Levitte <levitte@stacken.kth.se>]
2276 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
2277 [Richard Levitte <levitte@stacken.kth.se>]
2279 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
2280 from `int' to `unsigned int' because it's a length and initialized by
2281 EVP_DigestFinal() which expects an `unsigned int *'.
2282 [Richard Levitte <levitte@stacken.kth.se>]
2284 *) Don't hard-code path to Perl interpreter on shebang line of Configure
2285 script. Instead use the usual Shell->Perl transition trick.
2286 [Ralf S. Engelschall]
2288 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
2289 (in addition to RSA certificates) to match the behaviour of `openssl dsa
2290 -noout -modulus' as it's already the case for `openssl rsa -noout
2291 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
2292 currently the public key is printed (a decision which was already done by
2293 `openssl dsa -modulus' in the past) which serves a similar purpose.
2294 Additionally the NO_RSA no longer completely removes the whole -modulus
2295 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
2297 [Ralf S. Engelschall]
2299 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
2300 BIO. See the source (crypto/evp/bio_ok.c) for more info.
2301 [Arne Ansper <arne@ats.cyber.ee>]
2303 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
2304 to be added. Now both 'req' and 'ca' can use new objects defined in the
2308 *) Add cool BIO that does syslog (or event log on NT).
2309 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
2311 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
2312 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
2313 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
2314 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
2317 *) Add preliminary config info for new extension code.
2320 *) Make RSA_NO_PADDING really use no padding.
2321 [Ulf Moeller <ulf@fitug.de>]
2323 *) Generate errors when private/public key check is done.
2326 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
2327 for some CRL extensions and new objects added.
2330 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
2331 key usage extension and fuller support for authority key id.
2334 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
2335 padding method for RSA, which is recommended for new applications in PKCS
2336 #1 v2.0 (RFC 2437, October 1998).
2337 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
2338 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
2339 against Bleichbacher's attack on RSA.
2340 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
2343 *) Updates to the new SSL compression code
2344 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2346 *) Fix so that the version number in the master secret, when passed
2347 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
2348 (because the server will not accept higher), that the version number
2349 is 0x03,0x01, not 0x03,0x00
2350 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2352 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
2353 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
2354 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
2357 *) Support for RAW extensions where an arbitrary extension can be
2358 created by including its DER encoding. See apps/openssl.cnf for
2362 *) Make sure latest Perl versions don't interpret some generated C array
2363 code as Perl array code in the crypto/err/err_genc.pl script.
2364 [Lars Weber <3weber@informatik.uni-hamburg.de>]
2366 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
2367 not many people have the assembler. Various Win32 compilation fixes and
2368 update to the INSTALL.W32 file with (hopefully) more accurate Win32
2372 *) Modify configure script 'Configure' to automatically create crypto/date.h
2373 file under Win32 and also build pem.h from pem.org. New script
2374 util/mkfiles.pl to create the MINFO file on environments that can't do a
2375 'make files': perl util/mkfiles.pl >MINFO should work.
2378 *) Major rework of DES function declarations, in the pursuit of correctness
2379 and purity. As a result, many evil casts evaporated, and some weirdness,
2380 too. You may find this causes warnings in your code. Zapping your evil
2381 casts will probably fix them. Mostly.
2384 *) Fix for a typo in asn1.h. Bug fix to object creation script
2385 obj_dat.pl. It considered a zero in an object definition to mean
2386 "end of object": none of the objects in objects.h have any zeros
2387 so it wasn't spotted.
2388 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2390 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2391 Masking (CBCM). In the absence of test vectors, the best I have been able
2392 to do is check that the decrypt undoes the encrypt, so far. Send me test
2393 vectors if you have them.
2396 *) Correct calculation of key length for export ciphers (too much space was
2397 allocated for null ciphers). This has not been tested!
2400 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2401 message is now correct (it understands "crypto" and "ssl" on its
2402 command line). There is also now an "update" option. This will update
2403 the util/ssleay.num and util/libeay.num files with any new functions.
2405 perl util/mkdef.pl crypto ssl update
2406 it will update them.
2409 *) Overhauled the Perl interface (perl/*):
2410 - ported BN stuff to OpenSSL's different BN library
2411 - made the perl/ source tree CVS-aware
2412 - renamed the package from SSLeay to OpenSSL (the files still contain
2413 their history because I've copied them in the repository)
2414 - removed obsolete files (the test scripts will be replaced
2415 by better Test::Harness variants in the future)
2416 [Ralf S. Engelschall]
2418 *) First cut for a very conservative source tree cleanup:
2419 1. merge various obsolete readme texts into doc/ssleay.txt
2420 where we collect the old documents and readme texts.
2421 2. remove the first part of files where I'm already sure that we no
2422 longer need them because of three reasons: either they are just temporary
2423 files which were left by Eric or they are preserved original files where
2424 I've verified that the diff is also available in the CVS via "cvs diff
2425 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2426 the crypto/md/ stuff).
2427 [Ralf S. Engelschall]
2429 *) More extension code. Incomplete support for subject and issuer alt
2430 name, issuer and authority key id. Change the i2v function parameters
2431 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2432 what that's for :-) Fix to ASN1 macro which messed up
2433 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2436 *) Preliminary support for ENUMERATED type. This is largely copied from the
2440 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2441 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2443 *) Make sure `make rehash' target really finds the `openssl' program.
2444 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2446 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2447 like to hear about it if this slows down other processors.
2450 *) Add CygWin32 platform information to Configure script.
2451 [Alan Batie <batie@aahz.jf.intel.com>]
2453 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2454 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2456 *) New program nseq to manipulate netscape certificate sequences
2459 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2463 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2464 but the BN code had some problems that would cause failures when
2465 doing certificate verification and some other functions.
2466 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2468 *) Add ASN1 and PEM code to support netscape certificate sequences.
2471 *) Add ASN1 and PEM code to support netscape certificate sequences.
2474 *) Add several PKIX and private extended key usage OIDs.
2477 *) Modify the 'ca' program to handle the new extension code. Modify
2478 openssl.cnf for new extension format, add comments.
2481 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2482 and add a sample to openssl.cnf so req -x509 now adds appropriate
2486 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2487 error code, add initial support to X509_print() and x509 application.
2490 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2491 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2492 stuff is currently isolated and isn't even compiled yet.
2495 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2496 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2497 Removed the versions check from X509 routines when loading extensions:
2498 this allows certain broken certificates that don't set the version
2499 properly to be processed.
2502 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2503 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2504 can still be regenerated with "make depend".
2507 *) Spelling mistake in C version of CAST-128.
2508 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2510 *) Changes to the error generation code. The perl script err-code.pl
2511 now reads in the old error codes and retains the old numbers, only
2512 adding new ones if necessary. It also only changes the .err files if new
2513 codes are added. The makefiles have been modified to only insert errors
2514 when needed (to avoid needlessly modifying header files). This is done
2515 by only inserting errors if the .err file is newer than the auto generated
2516 C file. To rebuild all the error codes from scratch (the old behaviour)
2517 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2518 or delete all the .err files.
2521 *) CAST-128 was incorrectly implemented for short keys. The C version has
2522 been fixed, but is untested. The assembler versions are also fixed, but
2523 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2524 to regenerate it if needed.
2525 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2526 Hagino <itojun@kame.net>]
2528 *) File was opened incorrectly in randfile.c.
2529 [Ulf Möller <ulf@fitug.de>]
2531 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2532 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2533 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2534 al: it's just almost always a UTCTime. Note this patch adds new error
2535 codes so do a "make errors" if there are problems.
2538 *) Correct Linux 1 recognition in config.
2539 [Ulf Möller <ulf@fitug.de>]
2541 *) Remove pointless MD5 hash when using DSA keys in ca.
2542 [Anonymous <nobody@replay.com>]
2544 *) Generate an error if given an empty string as a cert directory. Also
2545 generate an error if handed NULL (previously returned 0 to indicate an
2546 error, but didn't set one).
2547 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
2549 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
2552 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
2553 parameters. This was causing a warning which killed off the Win32 compile.
2556 *) Remove C++ style comments from crypto/bn/bn_local.h.
2557 [Neil Costigan <neil.costigan@celocom.com>]
2559 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
2560 based on a text string, looking up short and long names and finally
2561 "dot" format. The "dot" format stuff didn't work. Added new function
2562 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
2563 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
2564 OID is not part of the table.
2567 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
2568 X509_LOOKUP_by_alias().
2571 *) Sort openssl functions by name.
2574 *) Get the gendsa program working (hopefully) and add it to app list. Remove
2575 encryption from sample DSA keys (in case anyone is interested the password
2579 *) Make _all_ *_free functions accept a NULL pointer.
2580 [Frans Heymans <fheymans@isaserver.be>]
2582 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
2584 [Anonymous <nobody@replay.com>]
2586 *) s_server should send the CAfile as acceptable CAs, not its own cert.
2587 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2589 *) Don't blow it for numeric -newkey arguments to apps/req.
2590 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2592 *) Temp key "for export" tests were wrong in s3_srvr.c.
2593 [Anonymous <nobody@replay.com>]
2595 *) Add prototype for temp key callback functions
2596 SSL_CTX_set_tmp_{rsa,dh}_callback().
2599 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
2600 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
2603 *) X509_name_add_entry() freed the wrong thing after an error.
2604 [Arne Ansper <arne@ats.cyber.ee>]
2606 *) rsa_eay.c would attempt to free a NULL context.
2607 [Arne Ansper <arne@ats.cyber.ee>]
2609 *) BIO_s_socket() had a broken should_retry() on Windoze.
2610 [Arne Ansper <arne@ats.cyber.ee>]
2612 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
2613 [Arne Ansper <arne@ats.cyber.ee>]
2615 *) Make sure the already existing X509_STORE->depth variable is initialized
2616 in X509_STORE_new(), but document the fact that this variable is still
2617 unused in the certificate verification process.
2618 [Ralf S. Engelschall]
2620 *) Fix the various library and apps files to free up pkeys obtained from
2621 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
2624 *) Fix reference counting in X509_PUBKEY_get(). This makes
2625 demos/maurice/example2.c work, amongst others, probably.
2626 [Steve Henson and Ben Laurie]
2628 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
2629 `openssl' and second, the shortcut symlinks for the `openssl <command>'
2630 are no longer created. This way we have a single and consistent command
2631 line interface `openssl <command>', similar to `cvs <command>'.
2632 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
2634 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
2635 BIT STRING wrapper always have zero unused bits.
2638 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
2641 *) Make the top-level INSTALL documentation easier to understand.
2644 *) Makefiles updated to exit if an error occurs in a sub-directory
2645 make (including if user presses ^C) [Paul Sutton]
2647 *) Make Montgomery context stuff explicit in RSA data structure.
2650 *) Fix build order of pem and err to allow for generated pem.h.
2653 *) Fix renumbering bug in X509_NAME_delete_entry().
2656 *) Enhanced the err-ins.pl script so it makes the error library number
2657 global and can add a library name. This is needed for external ASN1 and
2658 other error libraries.
2661 *) Fixed sk_insert which never worked properly.
2664 *) Fix ASN1 macros so they can handle indefinite length construted
2665 EXPLICIT tags. Some non standard certificates use these: they can now
2669 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
2670 into a single doc/ssleay.txt bundle. This way the information is still
2671 preserved but no longer messes up this directory. Now it's new room for
2672 the new set of documenation files.
2673 [Ralf S. Engelschall]
2675 *) SETs were incorrectly DER encoded. This was a major pain, because they
2676 shared code with SEQUENCEs, which aren't coded the same. This means that
2677 almost everything to do with SETs or SEQUENCEs has either changed name or
2678 number of arguments.
2679 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
2681 *) Fix test data to work with the above.
2684 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
2685 was already fixed by Eric for 0.9.1 it seems.
2686 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
2688 *) Autodetect FreeBSD3.
2691 *) Fix various bugs in Configure. This affects the following platforms:
2695 unixware-2.0-pentium
2699 *) Eliminate generated files from CVS. Reorder tests to regenerate files
2700 before they are needed.
2703 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
2707 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
2709 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
2710 changed SSLeay to OpenSSL in version strings.
2711 [Ralf S. Engelschall]
2713 *) Some fixups to the top-level documents.
2716 *) Fixed the nasty bug where rsaref.h was not found under compile-time
2717 because the symlink to include/ was missing.
2718 [Ralf S. Engelschall]
2720 *) Incorporated the popular no-RSA/DSA-only patches
2721 which allow to compile a RSA-free SSLeay.
2722 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
2724 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
2725 when "ssleay" is still not found.
2726 [Ralf S. Engelschall]
2728 *) Added more platforms to Configure: Cray T3E, HPUX 11,
2729 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
2731 *) Updated the README file.
2732 [Ralf S. Engelschall]
2734 *) Added various .cvsignore files in the CVS repository subdirs
2735 to make a "cvs update" really silent.
2736 [Ralf S. Engelschall]
2738 *) Recompiled the error-definition header files and added
2739 missing symbols to the Win32 linker tables.
2740 [Ralf S. Engelschall]
2742 *) Cleaned up the top-level documents;
2743 o new files: CHANGES and LICENSE
2744 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
2745 o merged COPYRIGHT into LICENSE
2746 o removed obsolete TODO file
2747 o renamed MICROSOFT to INSTALL.W32
2748 [Ralf S. Engelschall]
2750 *) Removed dummy files from the 0.9.1b source tree:
2751 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
2752 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
2753 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
2754 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
2755 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
2756 [Ralf S. Engelschall]
2758 *) Added various platform portability fixes.
2761 *) The Genesis of the OpenSSL rpject:
2762 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
2763 Young and Tim J. Hudson created while they were working for C2Net until
2765 [The OpenSSL Project]
2768 Changes between 0.9.0b and 0.9.1b [not released]
2770 *) Updated a few CA certificates under certs/
2773 *) Changed some BIGNUM api stuff.
2776 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
2777 DGUX x86, Linux Alpha, etc.
2780 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
2781 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
2785 *) Add -strparse option to asn1pars program which parses nested
2787 [Dr Stephen Henson <shenson@bigfoot.com>]
2789 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
2792 *) DSA fix for "ca" program.
2795 *) Added "-genkey" option to "dsaparam" program.
2798 *) Added RIPE MD160 (rmd160) message digest.
2801 *) Added -a (all) option to "ssleay version" command.
2804 *) Added PLATFORM define which is the id given to Configure.
2807 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
2810 *) Extended the ASN.1 parser routines.
2813 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
2816 *) Added a BN_CTX to the BN library.
2819 *) Fixed the weak key values in DES library
2822 *) Changed API in EVP library for cipher aliases.
2825 *) Added support for RC2/64bit cipher.
2828 *) Converted the lhash library to the crypto/mem.c functions.
2831 *) Added more recognized ASN.1 object ids.
2834 *) Added more RSA padding checks for SSL/TLS.
2837 *) Added BIO proxy/filter functionality.
2840 *) Added extra_certs to SSL_CTX which can be used
2841 send extra CA certificates to the client in the CA cert chain sending
2842 process. It can be configured with SSL_CTX_add_extra_chain_cert().
2845 *) Now Fortezza is denied in the authentication phase because
2846 this is key exchange mechanism is not supported by SSLeay at all.
2849 *) Additional PKCS1 checks.
2852 *) Support the string "TLSv1" for all TLS v1 ciphers.
2855 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
2856 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
2859 *) Fixed a few memory leaks.
2862 *) Fixed various code and comment typos.
2865 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
2866 bytes sent in the client random.
2867 [Edward Bishop <ebishop@spyglass.com>]