5 Changes between 0.9.4 and 0.9.5 [xx XXX 1999]
7 *) RAND_load_file(..., -1) reads the complete file and uses it to
8 seed the PRNG (previously an explicit byte count was required).
9 The bytes obtained by stat() are no longer counted for the
11 [Bodo Möller, Ulf Möller]
13 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
14 used (char *) instead of (void *) and had casts all over the place.
17 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
20 *) Retain source code compatibility for BN_prime_checks macro.
23 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
24 DH_CHECK_P_NOT_SAFE_PRIME.
25 (Check if this is true? OpenPGP calls them "strong".)
28 *) Merge the functionality of "dh" and "gendh" programs into a new program
29 "dhparam". The old programs are retained for now but will handle DH keys
30 (instead of parameters) in future.
33 *) Make the ciphers, s_server and s_client programs check the return values
34 when a new cipher list is set.
37 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
38 ciphers. Before when the 56bit ciphers were enabled the sorting was
41 The syntax for the cipher sorting has been extended to support sorting by
42 cipher-strength (using the strength_bits hard coded in the tables).
43 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
45 Fix a bug in the cipher-command parser: when supplying a cipher command
46 string with an "undefined" symbol (neither command nor alphanumeric
47 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
50 Due to the strength-sorting extension, the code of the
51 ssl_create_cipher_list() function was completely rearranged. I hope that
52 the readability was also increased :-)
53 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
55 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
56 for the first serial number and places 2 in the serial number file. This
57 avoids problems when the root CA is created with serial number zero and
58 the first user certificate has the same issuer name and serial number
62 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
63 the new code. Add documentation for this stuff.
66 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
67 X509_*() to X509at_*() on the grounds that they don't handle X509
68 structures and behave in an analagous way to the X509v3 functions:
69 they shouldn't be called directly but wrapper functions should be used
72 So we also now have some wrapper functions that call the X509at functions
73 when passed certificate requests. (TO DO: similar things can be done with
74 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
75 things. Some of these need some d2i or i2d and print functionality
76 because they handle more complex structures.)
79 *) Add missing #ifndefs that caused missing symbols when building libssl
80 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
82 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
84 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
85 has a return value which indicates the quality of the random data
86 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
87 error queue. New function RAND_pseudo_bytes() generates output that is
88 guaranteed to be unique but not unpredictable.
91 *) Do more iterations of Rabin-Miller probable prime test (specifically,
92 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
93 instead of only 2 for all lengths; see BN_prime_checks_size definition
94 in crypto/bn/bn_prime.c for the complete table). This guarantees a
95 false-positive rate of at most 2^-80 (actually less because we are
96 additionally doing trial division) for random input.
99 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
102 *) New function X509_CTX_rget_chain(), this returns the chain
103 from an X509_CTX structure with a dup of the stack and all
104 the X509 reference counts upped: so the stack will exist
105 after X509_CTX_cleanup() has been called. Modify pkcs12.c
108 Also make SSL_SESSION_print() print out the verify return
112 *) Add manpage for the pkcs12 command. Also change the default
113 behaviour so MAC iteration counts are used unless the new
114 -nomaciter option is used. This improves file security and
115 only older versions of MSIE (4.0 for example) need it.
118 *) Honor the no-xxx Configure options when creating .DEF files.
121 *) Add PKCS#10 attributes to field table: challengePassword,
122 unstructuredName and unstructuredAddress. These are taken from
123 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
124 international characters are used.
126 More changes to X509_ATTRIBUTE code: allow the setting of types
127 based on strings. Remove the 'loc' parameter when adding
128 attributes because these will be a SET OF encoding which is sorted
132 *) Initial changes to the 'req' utility to allow request generation
133 automation. This will allow an application to just generate a template
134 file containing all the field values and have req construct the
137 Initial support for X509_ATTRIBUTE handling. Stacks of these are
138 used all over the place including certificate requests and PKCS#7
139 structures. They are currently handled manually where necessary with
140 some primitive wrappers for PKCS#7. The new functions behave in a
141 manner analogous to the X509 extension functions: they allow
142 attributes to be looked up by NID and added.
144 Later something similar to the X509V3 code would be desirable to
145 automatically handle the encoding, decoding and printing of the
146 more complex types. The string types like challengePassword can
147 be handled by the string table functions.
149 Also modified the multi byte string table handling. Now there is
150 a 'global mask' which masks out certain types. The table itself
151 can use the flag STABLE_NO_MASK to ignore the mask setting: this
152 is useful when for example there is only one permissible type
153 (as in countryName) and using the mask might result in no valid
157 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
158 SSL_get_peer_finished to allow applications to obtain the latest
159 Finished messages sent to the peer or expected from the peer,
160 respectively. (SSL_get_peer_finished is usually the Finished message
161 actually received from the peer, otherwise the protocol will be aborted.)
163 As the Finished message are message digests of the complete handshake
164 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
165 be used for external authentication procedures when the authentication
166 provided by SSL/TLS is not desired or is not enough.
169 *) Enhanced support for Alpha Linux is added. Now ./config checks if
170 the host supports BWX extension and if Compaq C is present on the
171 $PATH. Just exploiting of the BWX extension results in 20-30%
172 performance kick for some algorithms, e.g. DES and RC4 to mention
173 a couple. Compaq C in turn generates ~20% faster code for MD5 and
177 *) Add support for MS "fast SGC". This is arguably a violation of the
178 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
179 weak crypto and after checking the certificate is SGC a second one
180 with strong crypto. MS SGC stops the first handshake after receiving
181 the server certificate message and sends a second client hello. Since
182 a server will typically do all the time consuming operations before
183 expecting any further messages from the client (server key exchange
184 is the most expensive) there is little difference between the two.
186 To get OpenSSL to support MS SGC we have to permit a second client
187 hello message after we have sent server done. In addition we have to
188 reset the MAC if we do get this second client hello and include the
192 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
193 if a DER encoded private key is RSA or DSA traditional format. Changed
194 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
195 format DER encoded private key. Newer code should use PKCS#8 format which
196 has the key type encoded in the ASN1 structure. Added DER private key
197 support to pkcs8 application.
200 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
201 ciphersuites has been selected (as required by the SSL 3/TLS 1
202 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
203 is set, we interpret this as a request to violate the specification
204 (the worst that can happen is a handshake failure, and 'correct'
205 behaviour would result in a handshake failure anyway).
208 *) In SSL_CTX_add_session, take into account that there might be multiple
209 SSL_SESSION structures with the same session ID (e.g. when two threads
210 concurrently obtain them from an external cache).
211 The internal cache can handle only one SSL_SESSION with a given ID,
212 so if there's a conflict, we now throw out the old one to achieve
216 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
217 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
218 some routines that use cipher OIDs: some ciphers do not have OIDs
219 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
223 *) Simplify the trust setting structure and code. Now we just have
224 two sequences of OIDs for trusted and rejected settings. These will
225 typically have values the same as the extended key usage extension
226 and any application specific purposes.
228 The trust checking code now has a default behaviour: it will just
229 check for an object with the same NID as the passed id. Functions can
230 be provided to override either the default behaviour or the behaviour
231 for a given id. SSL client, server and email already have functions
232 in place for compatibility: they check the NID and also return "trusted"
233 if the certificate is self signed.
236 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
237 traditional format into an EVP_PKEY structure.
240 *) Add a password callback function PEM_cb() which either prompts for
241 a password if usr_data is NULL or otherwise assumes it is a null
242 terminated password. Allow passwords to be passed on command line
243 environment or config files in a few more utilities.
246 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
247 keys. Add some short names for PKCS#8 PBE algorithms and allow them
248 to be specified on the command line for the pkcs8 and pkcs12 utilities.
249 Update documentation.
252 *) Support for ASN1 "NULL" type. This could be handled before by using
253 ASN1_TYPE but there wasn't any function that would try to read a NULL
254 and produce an error if it couldn't. For compatibility we also have
255 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
256 don't allocate anything because they don't need to.
259 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
261 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
263 *) Rebuild of the memory allocation routines used by OpenSSL code and
264 possibly others as well. The purpose is to make an interface that
265 provide hooks so anyone can build a separate set of allocation and
266 deallocation routines to be used by OpenSSL, for example memory
267 pool implementations, or something else, which was previously hard
268 since Malloc(), Realloc() and Free() were defined as macros having
269 the values malloc, realloc and free, respectively (except for Win32
270 compilations). The same is provided for memory debugging code.
271 OpenSSL already comes with functionality to find memory leaks, but
272 this gives people a chance to debug other memory problems.
274 With these changes, a new set of functions and macros have appeared:
276 CRYPTO_set_mem_debug_functions() [F]
277 CRYPTO_get_mem_debug_functions() [F]
278 CRYPTO_dbg_set_options() [F]
279 CRYPTO_dbg_get_options() [F]
280 CRYPTO_malloc_debug_init() [M]
282 The memory debug functions are NULL by default, unless the library
283 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
284 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
285 gives the standard debugging functions that come with OpenSSL) or
286 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
287 provided by the library user) must be used. When the standard
288 debugging functions are used, CRYPTO_dbg_set_options can be used to
289 request additional information:
290 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
291 the CRYPTO_MDEBUG_xxx macro when compiling the library.
293 Also, things like CRYPTO_set_mem_functions will always give the
294 expected result (the new set of functions is used for allocation
295 and deallocation) at all times, regardless of platform and compiler
298 To finish it up, some functions that were never use in any other
299 way than through macros have a new API and new semantic:
305 All macros of value have retained their old syntax.
306 [Richard Levitte and Bodo Moeller]
308 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
309 ordering of SMIMECapabilities wasn't in "strength order" and there
310 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
314 *) Some ASN1 types with illegal zero length encoding (INTEGER,
315 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
316 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
318 *) Merge in my S/MIME library for OpenSSL. This provides a simple
319 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
320 functionality to handle multipart/signed properly) and a utility
321 called 'smime' to call all this stuff. This is based on code I
322 originally wrote for Celo who have kindly allowed it to be
326 *) Add variants des_set_key_checked and des_set_key_unchecked of
327 des_set_key (aka des_key_sched). Global variable des_check_key
328 decides which of these is called by des_set_key; this way
329 des_check_key behaves as it always did, but applications and
330 the library itself, which was buggy for des_check_key == 1,
331 have a cleaner way to pick the version they need.
334 *) New function PKCS12_newpass() which changes the password of a
338 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
339 dynamic mix. In both cases the ids can be used as an index into the
340 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
341 functions so they accept a list of the field values and the
342 application doesn't need to directly manipulate the X509_TRUST
346 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
350 *) Modify the way the V3 extension code looks up extensions. This now
351 works in a similar way to the object code: we have some "standard"
352 extensions in a static table which is searched with OBJ_bsearch()
353 and the application can add dynamic ones if needed. The file
354 crypto/x509v3/ext_dat.h now has the info: this file needs to be
355 updated whenever a new extension is added to the core code and kept
356 in ext_nid order. There is a simple program 'tabtest.c' which checks
357 this. New extensions are not added too often so this file can readily
358 be maintained manually.
360 There are two big advantages in doing things this way. The extensions
361 can be looked up immediately and no longer need to be "added" using
362 X509V3_add_standard_extensions(): this function now does nothing.
363 [Side note: I get *lots* of email saying the extension code doesn't
364 work because people forget to call this function]
365 Also no dynamic allocation is done unless new extensions are added:
366 so if we don't add custom extensions there is no need to call
367 X509V3_EXT_cleanup().
370 *) Modify enc utility's salting as follows: make salting the default. Add a
371 magic header, so unsalted files fail gracefully instead of just decrypting
372 to garbage. This is because not salting is a big security hole, so people
373 should be discouraged from doing it.
376 *) Fixes and enhancements to the 'x509' utility. It allowed a message
377 digest to be passed on the command line but it only used this
378 parameter when signing a certificate. Modified so all relevant
379 operations are affected by the digest parameter including the
380 -fingerprint and -x509toreq options. Also -x509toreq choked if a
381 DSA key was used because it didn't fix the digest.
384 *) Initial certificate chain verify code. Currently tests the untrusted
385 certificates for consistency with the verify purpose (which is set
386 when the X509_STORE_CTX structure is set up) and checks the pathlength.
388 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
389 this is because it will reject chains with invalid extensions whereas
390 every previous version of OpenSSL and SSLeay made no checks at all.
392 Trust code: checks the root CA for the relevant trust settings. Trust
393 settings have an initial value consistent with the verify purpose: e.g.
394 if the verify purpose is for SSL client use it expects the CA to be
395 trusted for SSL client use. However the default value can be changed to
396 permit custom trust settings: one example of this would be to only trust
397 certificates from a specific "secure" set of CAs.
399 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
400 which should be used for version portability: especially since the
401 verify structure is likely to change more often now.
403 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
404 to set them. If not set then assume SSL clients will verify SSL servers
407 Two new options to the verify program: -untrusted allows a set of
408 untrusted certificates to be passed in and -purpose which sets the
409 intended purpose of the certificate. If a purpose is set then the
410 new chain verify code is used to check extension consistency.
413 *) Support for the authority information access extension.
416 *) Modify RSA and DSA PEM read routines to transparently handle
417 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
418 public keys in a format compatible with certificate
419 SubjectPublicKeyInfo structures. Unfortunately there were already
420 functions called *_PublicKey_* which used various odd formats so
421 these are retained for compatibility: however the DSA variants were
422 never in a public release so they have been deleted. Changed dsa/rsa
423 utilities to handle the new format: note no releases ever handled public
424 keys so we should be OK.
426 The primary motivation for this change is to avoid the same fiasco
427 that dogs private keys: there are several incompatible private key
428 formats some of which are standard and some OpenSSL specific and
429 require various evil hacks to allow partial transparent handling and
430 even then it doesn't work with DER formats. Given the option anything
431 other than PKCS#8 should be dumped: but the other formats have to
432 stay in the name of compatibility.
434 With public keys and the benefit of hindsight one standard format
435 is used which works with EVP_PKEY, RSA or DSA structures: though
436 it clearly returns an error if you try to read the wrong kind of key.
438 Added a -pubkey option to the 'x509' utility to output the public key.
439 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*() and add
440 EVP_PKEY_rset_*() functions that do the same as the EVP_PKEY_assign_*()
441 except they up the reference count of the added key (they don't "swallow"
445 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
446 CRLs would fail if the file contained no certificates or no CRLs:
447 added a new function to read in both types and return the number
448 read: this means that if none are read it will be an error. The
449 DER versions of the certificate and CRL reader would always fail
450 because it isn't possible to mix certificates and CRLs in DER format
451 without choking one or the other routine. Changed this to just read
452 a certificate: this is the best we can do. Also modified the code
453 in apps/verify.c to take notice of return codes: it was previously
454 attempting to read in certificates from NULL pointers and ignoring
455 any errors: this is one reason why the cert and CRL reader seemed
456 to work. It doesn't check return codes from the default certificate
457 routines: these may well fail if the certificates aren't installed.
460 *) Code to support otherName option in GeneralName.
463 *) First update to verify code. Change the verify utility
464 so it warns if it is passed a self signed certificate:
465 for consistency with the normal behaviour. X509_verify
466 has been modified to it will now verify a self signed
467 certificate if *exactly* the same certificate appears
468 in the store: it was previously impossible to trust a
469 single self signed certificate. This means that:
470 openssl verify ss.pem
471 now gives a warning about a self signed certificate but
472 openssl verify -CAfile ss.pem ss.pem
476 *) For servers, store verify_result in SSL_SESSION data structure
477 (and add it to external session representation).
478 This is needed when client certificate verifications fails,
479 but an application-provided verification callback (set by
480 SSL_CTX_set_cert_verify_callback) allows accepting the session
481 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
482 but returns 1): When the session is reused, we have to set
483 ssl->verify_result to the appropriate error code to avoid
485 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
487 *) Fix a bug in the new PKCS#7 code: it didn't consider the
488 case in PKCS7_dataInit() where the signed PKCS7 structure
489 didn't contain any existing data because it was being created.
490 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
492 *) Add a salt to the key derivation routines in enc.c. This
493 forms the first 8 bytes of the encrypted file. Also add a
494 -S option to allow a salt to be input on the command line.
497 *) New function X509_cmp(). Oddly enough there wasn't a function
498 to compare two certificates. We do this by working out the SHA1
499 hash and comparing that. X509_cmp() will be needed by the trust
503 *) Correctly increment the reference count in the SSL_SESSION pointer
504 returned from SSL_get_session().
505 [Geoff Thorpe <geoff@eu.c2.net>]
507 *) Fix for 'req': it was adding a null to request attributes.
508 Also change the X509_LOOKUP and X509_INFO code to handle
509 certificate auxiliary information.
512 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
516 *) Add the possibility to add extra information to the memory leak
517 detecting output, to form tracebacks, showing from where each
518 allocation was originated: CRYPTO_push_info("constant string") adds
519 the string plus current file name and line number to a per-thread
520 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
521 is like calling CYRPTO_pop_info() until the stack is empty.
522 Also updated memory leak detection code to be multi-thread-safe.
525 *) Add options -text and -noout to pkcs7 utility and delete the
526 encryption options which never did anything. Update docs.
529 *) Add options to some of the utilities to allow the pass phrase
530 to be included on either the command line (not recommended on
531 OSes like Unix) or read from the environment. Update the
532 manpages and fix a few bugs.
535 *) Add a few manpages for some of the openssl commands.
538 *) Fix the -revoke option in ca. It was freeing up memory twice,
539 leaking and not finding already revoked certificates.
542 *) Extensive changes to support certificate auxiliary information.
543 This involves the use of X509_CERT_AUX structure and X509_AUX
544 functions. An X509_AUX function such as PEM_read_X509_AUX()
545 can still read in a certificate file in the usual way but it
546 will also read in any additional "auxiliary information". By
547 doing things this way a fair degree of compatibility can be
548 retained: existing certificates can have this information added
549 using the new 'x509' options.
551 Current auxiliary information includes an "alias" and some trust
552 settings. The trust settings will ultimately be used in enhanced
553 certificate chain verification routines: currently a certificate
554 can only be trusted if it is self signed and then it is trusted
558 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
559 The problem was that one of the replacement routines had not been working
560 since SSLeay releases. For now the offending routine has been replaced
561 with non-optimised assembler. Even so, this now gives around 95%
562 performance improvement for 1024 bit RSA signs.
565 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
566 handling. Most clients have the effective key size in bits equal to
567 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
568 A few however don't do this and instead use the size of the decrypted key
569 to determine the RC2 key length and the AlgorithmIdentifier to determine
570 the effective key length. In this case the effective key length can still
571 be 40 bits but the key length can be 168 bits for example. This is fixed
572 by manually forcing an RC2 key into the EVP_PKEY structure because the
573 EVP code can't currently handle unusual RC2 key sizes: it always assumes
574 the key length and effective key length are equal.
577 *) Add a bunch of functions that should simplify the creation of
578 X509_NAME structures. Now you should be able to do:
579 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
580 and have it automatically work out the correct field type and fill in
581 the structures. The more adventurous can try:
582 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
583 and it will (hopefully) work out the correct multibyte encoding.
586 *) Change the 'req' utility to use the new field handling and multibyte
587 copy routines. Before the DN field creation was handled in an ad hoc
588 way in req, ca, and x509 which was rather broken and didn't support
589 BMPStrings or UTF8Strings. Since some software doesn't implement
590 BMPStrings or UTF8Strings yet, they can be enabled using the config file
591 using the dirstring_type option. See the new comment in the default
592 openssl.cnf for more info.
595 *) Make crypto/rand/md_rand.c more robust:
596 - Assure unique random numbers after fork().
597 - Make sure that concurrent threads access the global counter and
598 md serializably so that we never lose entropy in them
599 or use exactly the same state in multiple threads.
600 Access to the large state is not always serializable because
601 the additional locking could be a performance killer, and
602 md should be large enough anyway.
605 *) New file apps/app_rand.c with commonly needed functionality
606 for handling the random seed file.
608 Use the random seed file in some applications that previously did not:
610 dsaparam -genkey (which also ignored its '-rand' option),
614 Except on systems with /dev/urandom, it is crucial to have a random
615 seed file at least for key creation, DSA signing, and for DH exchanges;
616 for RSA signatures we could do without one.
618 gendh and gendsa (unlike genrsa) used to read only the first byte
619 of each file listed in the '-rand' option. The function as previously
620 found in genrsa is now in app_rand.c and is used by all programs
621 that support '-rand'.
624 *) In RAND_write_file, use mode 0600 for creating files;
625 don't just chmod when it may be too late.
628 *) Report an error from X509_STORE_load_locations
629 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
632 *) New function ASN1_mbstring_copy() this copies a string in either
633 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
634 into an ASN1_STRING type. A mask of permissible types is passed
635 and it chooses the "minimal" type to use or an error if not type
639 *) Add function equivalents to the various macros in asn1.h. The old
640 macros are retained with an M_ prefix. Code inside the library can
641 use the M_ macros. External code (including the openssl utility)
642 should *NOT* in order to be "shared library friendly".
645 *) Add various functions that can check a certificate's extensions
646 to see if it usable for various purposes such as SSL client,
647 server or S/MIME and CAs of these types. This is currently
648 VERY EXPERIMENTAL but will ultimately be used for certificate chain
649 verification. Also added a -purpose flag to x509 utility to
650 print out all the purposes.
653 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
657 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
658 for, obtain and decode and extension and obtain its critical flag.
659 This allows all the necessary extension code to be handled in a
660 single function call.
663 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
664 platforms. See crypto/rc4/rc4_enc.c for further details.
667 *) New -noout option to asn1parse. This causes no output to be produced
668 its main use is when combined with -strparse and -out to extract data
669 from a file (which may not be in ASN.1 format).
672 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
673 when producing the local key id.
674 [Richard Levitte <levitte@stacken.kth.se>]
676 *) New option -dhparam in s_server. This allows a DH parameter file to be
677 stated explicitly. If it is not stated then it tries the first server
678 certificate file. The previous behaviour hard coded the filename
682 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
683 a public key to be input or output. For example:
684 openssl rsa -in key.pem -pubout -out pubkey.pem
685 Also added necessary DSA public key functions to handle this.
688 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
689 in the message. This was handled by allowing
690 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
691 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
693 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
694 to the end of the strings whereas this didn't. This would cause problems
695 if strings read with d2i_ASN1_bytes() were later modified.
696 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
698 *) Fix for base64 decode bug. When a base64 bio reads only one line of
699 data and it contains EOF it will end up returning an error. This is
700 caused by input 46 bytes long. The cause is due to the way base64
701 BIOs find the start of base64 encoded data. They do this by trying a
702 trial decode on each line until they find one that works. When they
703 do a flag is set and it starts again knowing it can pass all the
704 data directly through the decoder. Unfortunately it doesn't reset
705 the context it uses. This means that if EOF is reached an attempt
706 is made to pass two EOFs through the context and this causes the
707 resulting error. This can also cause other problems as well. As is
708 usual with these problems it takes *ages* to find and the fix is
709 trivial: move one line.
710 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
712 *) Ugly workaround to get s_client and s_server working under Windows. The
713 old code wouldn't work because it needed to select() on sockets and the
714 tty (for keypresses and to see if data could be written). Win32 only
715 supports select() on sockets so we select() with a 1s timeout on the
716 sockets and then see if any characters are waiting to be read, if none
717 are present then we retry, we also assume we can always write data to
718 the tty. This isn't nice because the code then blocks until we've
719 received a complete line of data and it is effectively polling the
720 keyboard at 1s intervals: however it's quite a bit better than not
721 working at all :-) A dedicated Windows application might handle this
722 with an event loop for example.
725 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
726 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
727 will be called when RSA_sign() and RSA_verify() are used. This is useful
728 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
729 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
730 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
731 This necessitated the support of an extra signature type NID_md5_sha1
732 for SSL signatures and modifications to the SSL library to use it instead
733 of calling RSA_public_decrypt() and RSA_private_encrypt().
736 *) Add new -verify -CAfile and -CApath options to the crl program, these
737 will lookup a CRL issuers certificate and verify the signature in a
738 similar way to the verify program. Tidy up the crl program so it
739 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
740 less strict. It will now permit CRL extensions even if it is not
741 a V2 CRL: this will allow it to tolerate some broken CRLs.
744 *) Initialize all non-automatic variables each time one of the openssl
745 sub-programs is started (this is necessary as they may be started
746 multiple times from the "OpenSSL>" prompt).
747 [Lennart Bang, Bodo Moeller]
749 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
750 removing all other RSA functionality (this is what NO_RSA does). This
751 is so (for example) those in the US can disable those operations covered
752 by the RSA patent while allowing storage and parsing of RSA keys and RSA
756 *) Non-copying interface to BIO pairs.
757 (still largely untested)
760 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
761 ASCII string. This was handled independently in various places before.
764 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
765 UTF8 strings a character at a time.
768 *) Use client_version from client hello to select the protocol
769 (s23_srvr.c) and for RSA client key exchange verification
770 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
773 *) Add various utility functions to handle SPKACs, these were previously
774 handled by poking round in the structure internals. Added new function
775 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
776 print, verify and generate SPKACs. Based on an original idea from
777 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
780 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
783 *) Allow the config file extension section to be overwritten on the
784 command line. Based on an original idea from Massimiliano Pala
785 <madwolf@comune.modena.it>. The new option is called -extensions
786 and can be applied to ca, req and x509. Also -reqexts to override
787 the request extensions in req and -crlexts to override the crl extensions
791 *) Add new feature to the SPKAC handling in ca. Now you can include
792 the same field multiple times by preceding it by "XXXX." for example:
795 this is the same syntax as used in the req config file.
798 *) Allow certificate extensions to be added to certificate requests. These
799 are specified in a 'req_extensions' option of the req section of the
800 config file. They can be printed out with the -text option to req but
801 are otherwise ignored at present.
804 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
805 data read consists of only the final block it would not decrypted because
806 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
807 A misplaced 'break' also meant the decrypted final block might not be
808 copied until the next read.
811 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
812 a few extra parameters to the DH structure: these will be useful if
813 for example we want the value of 'q' or implement X9.42 DH.
816 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
817 provides hooks that allow the default DSA functions or functions on a
818 "per key" basis to be replaced. This allows hardware acceleration and
819 hardware key storage to be handled without major modification to the
820 library. Also added low level modexp hooks and CRYPTO_EX structure and
821 associated functions.
824 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
825 as "read only": it can't be written to and the buffer it points to will
826 not be freed. Reading from a read only BIO is much more efficient than
827 a normal memory BIO. This was added because there are several times when
828 an area of memory needs to be read from a BIO. The previous method was
829 to create a memory BIO and write the data to it, this results in two
830 copies of the data and an O(n^2) reading algorithm. There is a new
831 function BIO_new_mem_buf() which creates a read only memory BIO from
832 an area of memory. Also modified the PKCS#7 routines to use read only
836 *) Bugfix: ssl23_get_client_hello did not work properly when called in
837 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
838 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
839 but a retry condition occured while trying to read the rest.
842 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
843 NID_pkcs7_encrypted by default: this was wrong since this should almost
844 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
845 the encrypted data type: this is a more sensible place to put it and it
846 allows the PKCS#12 code to be tidied up that duplicated this
850 *) Changed obj_dat.pl script so it takes its input and output files on
851 the command line. This should avoid shell escape redirection problems
855 *) Initial support for certificate extension requests, these are included
856 in things like Xenroll certificate requests. Included functions to allow
857 extensions to be obtained and added.
860 *) -crlf option to s_client and s_server for sending newlines as
861 CRLF (as required by many protocols).
864 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
866 *) Install libRSAglue.a when OpenSSL is built with RSAref.
867 [Ralf S. Engelschall]
869 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
870 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
872 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
876 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
877 DH parameters/keys (q is lost during that conversion, but the resulting
878 DH parameters contain its length).
880 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
881 much faster than DH_generate_parameters (which creates parameters
882 where p = 2*q + 1), and also the smaller q makes DH computations
883 much more efficient (160-bit exponentiation instead of 1024-bit
884 exponentiation); so this provides a convenient way to support DHE
885 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
886 utter importance to use
887 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
889 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
890 when such DH parameters are used, because otherwise small subgroup
891 attacks may become possible!
894 *) Avoid memory leak in i2d_DHparams.
897 *) Allow the -k option to be used more than once in the enc program:
898 this allows the same encrypted message to be read by multiple recipients.
901 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
902 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
903 it will always use the numerical form of the OID, even if it has a short
907 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
908 method only got called if p,q,dmp1,dmq1,iqmp components were present,
909 otherwise bn_mod_exp was called. In the case of hardware keys for example
910 no private key components need be present and it might store extra data
911 in the RSA structure, which cannot be accessed from bn_mod_exp.
912 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
913 private key operations.
916 *) Added support for SPARC Linux.
919 *) pem_password_cb function type incompatibly changed from
920 typedef int pem_password_cb(char *buf, int size, int rwflag);
922 ....(char *buf, int size, int rwflag, void *userdata);
923 so that applications can pass data to their callbacks:
924 The PEM[_ASN1]_{read,write}... functions and macros now take an
925 additional void * argument, which is just handed through whenever
926 the password callback is called.
927 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
929 New function SSL_CTX_set_default_passwd_cb_userdata.
931 Compatibility note: As many C implementations push function arguments
932 onto the stack in reverse order, the new library version is likely to
933 interoperate with programs that have been compiled with the old
934 pem_password_cb definition (PEM_whatever takes some data that
935 happens to be on the stack as its last argument, and the callback
936 just ignores this garbage); but there is no guarantee whatsoever that
939 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
940 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
941 problems not only on Windows, but also on some Unix platforms.
942 To avoid problematic command lines, these definitions are now in an
943 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
944 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
947 *) MIPS III/IV assembler module is reimplemented.
950 *) More DES library cleanups: remove references to srand/rand and
951 delete an unused file.
954 *) Add support for the the free Netwide assembler (NASM) under Win32,
955 since not many people have MASM (ml) and it can be hard to obtain.
956 This is currently experimental but it seems to work OK and pass all
957 the tests. Check out INSTALL.W32 for info.
960 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
961 without temporary keys kept an extra copy of the server key,
962 and connections with temporary keys did not free everything in case
966 *) New function RSA_check_key and new openssl rsa option -check
967 for verifying the consistency of RSA keys.
968 [Ulf Moeller, Bodo Moeller]
970 *) Various changes to make Win32 compile work:
971 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
972 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
973 comparison" warnings.
974 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
977 *) Add a debugging option to PKCS#5 v2 key generation function: when
978 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
979 derived keys are printed to stderr.
982 *) Copy the flags in ASN1_STRING_dup().
983 [Roman E. Pavlov <pre@mo.msk.ru>]
985 *) The x509 application mishandled signing requests containing DSA
986 keys when the signing key was also DSA and the parameters didn't match.
988 It was supposed to omit the parameters when they matched the signing key:
989 the verifying software was then supposed to automatically use the CA's
990 parameters if they were absent from the end user certificate.
992 Omitting parameters is no longer recommended. The test was also
993 the wrong way round! This was probably due to unusual behaviour in
994 EVP_cmp_parameters() which returns 1 if the parameters match.
995 This meant that parameters were omitted when they *didn't* match and
996 the certificate was useless. Certificates signed with 'ca' didn't have
998 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1000 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1001 The interface is as follows:
1002 Applications can use
1003 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1004 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1005 "off" is now the default.
1006 The library internally uses
1007 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1008 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1009 to disable memory-checking temporarily.
1011 Some inconsistent states that previously were possible (and were
1012 even the default) are now avoided.
1014 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1015 with each memory chunk allocated; this is occasionally more helpful
1016 than just having a counter.
1018 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1020 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1024 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1025 which largely parallels "options", but is for changing API behaviour,
1026 whereas "options" are about protocol behaviour.
1027 Initial "mode" flags are:
1029 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1030 a single record has been written.
1031 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1032 retries use the same buffer location.
1033 (But all of the contents must be
1037 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1040 *) Fix problems with no-hmac etc.
1041 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1043 *) New functions RSA_get_default_method(), RSA_set_method() and
1044 RSA_get_method(). These allows replacement of RSA_METHODs without having
1045 to mess around with the internals of an RSA structure.
1048 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1049 Also really enable memory leak checks in openssl.c and in some
1051 [Chad C. Mulligan, Bodo Moeller]
1053 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1054 up the length of negative integers. This has now been simplified to just
1055 store the length when it is first determined and use it later, rather
1056 than trying to keep track of where data is copied and updating it to
1058 [Steve Henson, reported by Brien Wheeler
1059 <bwheeler@authentica-security.com>]
1061 *) Add a new function PKCS7_signatureVerify. This allows the verification
1062 of a PKCS#7 signature but with the signing certificate passed to the
1063 function itself. This contrasts with PKCS7_dataVerify which assumes the
1064 certificate is present in the PKCS#7 structure. This isn't always the
1065 case: certificates can be omitted from a PKCS#7 structure and be
1066 distributed by "out of band" means (such as a certificate database).
1069 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1070 function prototypes in pem.h, also change util/mkdef.pl to add the
1071 necessary function names.
1074 *) mk1mf.pl (used by Windows builds) did not properly read the
1075 options set by Configure in the top level Makefile, and Configure
1076 was not even able to write more than one option correctly.
1077 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1080 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1081 file to be loaded from a BIO or FILE pointer. The BIO version will
1082 for example allow memory BIOs to contain config info.
1085 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1086 Whoever hopes to achieve shared-library compatibility across versions
1087 must use this, not the compile-time macro.
1088 (Exercise 0.9.4: Which is the minimum library version required by
1090 Note: All this applies only to multi-threaded programs, others don't
1094 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1095 through a BIO pair triggered the default case, i.e.
1096 SSLerr(...,SSL_R_UNKNOWN_STATE).
1099 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1100 can use the SSL library even if none of the specific BIOs is
1104 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1105 for the encoded length.
1106 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1108 *) Add initial documentation of the X509V3 functions.
1111 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1112 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1113 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1114 secure PKCS#8 private key format with a high iteration count.
1117 *) Fix determination of Perl interpreter: A perl or perl5
1118 _directory_ in $PATH was also accepted as the interpreter.
1119 [Ralf S. Engelschall]
1121 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1122 wrong with it but it was very old and did things like calling
1123 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1127 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1128 to use the new extension code.
1131 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1132 with macros. This should make it easier to change their form, add extra
1133 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1137 *) Add to configuration table a new entry that can specify an alternative
1138 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1139 according to Mark Crispin <MRC@Panda.COM>.
1143 *) DES CBC did not update the IV. Weird.
1146 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1147 Changing the behaviour of the former might break existing programs --
1148 where IV updating is needed, des_ncbc_encrypt can be used.
1151 *) When bntest is run from "make test" it drives bc to check its
1152 calculations, as well as internally checking them. If an internal check
1153 fails, it needs to cause bc to give a non-zero result or make test carries
1154 on without noticing the failure. Fixed.
1157 *) DES library cleanups.
1160 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
1161 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
1162 ciphers. NOTE: although the key derivation function has been verified
1163 against some published test vectors it has not been extensively tested
1164 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
1168 *) Instead of "mkdir -p", which is not fully portable, use new
1169 Perl script "util/mkdir-p.pl".
1172 *) Rewrite the way password based encryption (PBE) is handled. It used to
1173 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
1174 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
1175 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
1176 the 'parameter' field of the AlgorithmIdentifier is passed to the
1177 underlying key generation function so it must do its own ASN1 parsing.
1178 This has also changed the EVP_PBE_CipherInit() function which now has a
1179 'parameter' argument instead of literal salt and iteration count values
1180 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
1183 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
1184 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
1185 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
1186 KEY" because this clashed with PKCS#8 unencrypted string. Since this
1187 value was just used as a "magic string" and not used directly its
1188 value doesn't matter.
1191 *) Introduce some semblance of const correctness to BN. Shame C doesn't
1195 *) "linux-sparc64" configuration (ultrapenguin).
1196 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
1197 "linux-sparc" configuration.
1198 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
1200 *) config now generates no-xxx options for missing ciphers.
1203 *) Support the EBCDIC character set (work in progress).
1204 File ebcdic.c not yet included because it has a different license.
1205 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1207 *) Support BS2000/OSD-POSIX.
1208 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1210 *) Make callbacks for key generation use void * instead of char *.
1213 *) Make S/MIME samples compile (not yet tested).
1216 *) Additional typesafe stacks.
1219 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
1223 Changes between 0.9.3 and 0.9.3a [29 May 1999]
1225 *) New configuration variant "sco5-gcc".
1227 *) Updated some demos.
1228 [Sean O Riordain, Wade Scholine]
1230 *) Add missing BIO_free at exit of pkcs12 application.
1233 *) Fix memory leak in conf.c.
1236 *) Updates for Win32 to assembler version of MD5.
1239 *) Set #! path to perl in apps/der_chop to where we found it
1240 instead of using a fixed path.
1243 *) SHA library changes for irix64-mips4-cc.
1246 *) Improvements for VMS support.
1250 Changes between 0.9.2b and 0.9.3 [24 May 1999]
1252 *) Bignum library bug fix. IRIX 6 passes "make test" now!
1253 This also avoids the problems with SC4.2 and unpatched SC5.
1254 [Andy Polyakov <appro@fy.chalmers.se>]
1256 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
1257 These are required because of the typesafe stack would otherwise break
1258 existing code. If old code used a structure member which used to be STACK
1259 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
1260 sk_num or sk_value it would produce an error because the num, data members
1261 are not present in STACK_OF. Now it just produces a warning. sk_set
1262 replaces the old method of assigning a value to sk_value
1263 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
1264 that does this will no longer work (and should use sk_set instead) but
1265 this could be regarded as a "questionable" behaviour anyway.
1268 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
1269 correctly handle encrypted S/MIME data.
1272 *) Change type of various DES function arguments from des_cblock
1273 (which means, in function argument declarations, pointer to char)
1274 to des_cblock * (meaning pointer to array with 8 char elements),
1275 which allows the compiler to do more typechecking; it was like
1276 that back in SSLeay, but with lots of ugly casts.
1278 Introduce new type const_des_cblock.
1281 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
1282 problems: find RecipientInfo structure that matches recipient certificate
1283 and initialise the ASN1 structures properly based on passed cipher.
1286 *) Belatedly make the BN tests actually check the results.
1289 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
1290 to and from BNs: it was completely broken. New compilation option
1291 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
1292 key elements as negative integers.
1295 *) Reorganize and speed up MD5.
1296 [Andy Polyakov <appro@fy.chalmers.se>]
1299 [Richard Levitte <richard@levitte.org>]
1301 *) New option -out to asn1parse to allow the parsed structure to be
1302 output to a file. This is most useful when combined with the -strparse
1303 option to examine the output of things like OCTET STRINGS.
1306 *) Make SSL library a little more fool-proof by not requiring any longer
1307 that SSL_set_{accept,connect}_state be called before
1308 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
1309 in many applications because usually everything *appeared* to work as
1310 intended anyway -- now it really works as intended).
1313 *) Move openssl.cnf out of lib/.
1316 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
1317 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
1318 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
1319 [Ralf S. Engelschall]
1321 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
1322 handle PKCS#7 enveloped data properly.
1323 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
1325 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
1326 copying pointers. The cert_st handling is changed by this in
1327 various ways (and thus what used to be known as ctx->default_cert
1328 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
1329 any longer when s->cert does not give us what we need).
1330 ssl_cert_instantiate becomes obsolete by this change.
1331 As soon as we've got the new code right (possibly it already is?),
1332 we have solved a couple of bugs of the earlier code where s->cert
1333 was used as if it could not have been shared with other SSL structures.
1335 Note that using the SSL API in certain dirty ways now will result
1336 in different behaviour than observed with earlier library versions:
1337 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
1338 does not influence s as it used to.
1340 In order to clean up things more thoroughly, inside SSL_SESSION
1341 we don't use CERT any longer, but a new structure SESS_CERT
1342 that holds per-session data (if available); currently, this is
1343 the peer's certificate chain and, for clients, the server's certificate
1344 and temporary key. CERT holds only those values that can have
1345 meaningful defaults in an SSL_CTX.
1348 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
1349 from the internal representation. Various PKCS#7 fixes: remove some
1350 evil casts and set the enc_dig_alg field properly based on the signing
1354 *) Allow PKCS#12 password to be set from the command line or the
1355 environment. Let 'ca' get its config file name from the environment
1356 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
1360 *) Allow certificate policies extension to use an IA5STRING for the
1361 organization field. This is contrary to the PKIX definition but
1362 VeriSign uses it and IE5 only recognises this form. Document 'x509'
1366 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
1367 without disallowing inline assembler and the like for non-pedantic builds.
1370 *) Support Borland C++ builder.
1371 [Janez Jere <jj@void.si>, modified by Ulf Möller]
1376 *) SHA-1 cleanups and performance enhancements.
1377 [Andy Polyakov <appro@fy.chalmers.se>]
1379 *) Sparc v8plus assembler for the bignum library.
1380 [Andy Polyakov <appro@fy.chalmers.se>]
1382 *) Accept any -xxx and +xxx compiler options in Configure.
1385 *) Update HPUX configuration.
1388 *) Add missing sk_<type>_unshift() function to safestack.h
1389 [Ralf S. Engelschall]
1391 *) New function SSL_CTX_use_certificate_chain_file that sets the
1392 "extra_cert"s in addition to the certificate. (This makes sense
1393 only for "PEM" format files, as chains as a whole are not
1397 *) Support verify_depth from the SSL API.
1398 x509_vfy.c had what can be considered an off-by-one-error:
1399 Its depth (which was not part of the external interface)
1400 was actually counting the number of certificates in a chain;
1401 now it really counts the depth.
1404 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
1405 instead of X509err, which often resulted in confusing error
1406 messages since the error codes are not globally unique
1407 (e.g. an alleged error in ssl3_accept when a certificate
1408 didn't match the private key).
1410 *) New function SSL_CTX_set_session_id_context that allows to set a default
1411 value (so that you don't need SSL_set_session_id_context for each
1412 connection using the SSL_CTX).
1415 *) OAEP decoding bug fix.
1418 *) Support INSTALL_PREFIX for package builders, as proposed by
1422 *) New Configure options "threads" and "no-threads". For systems
1423 where the proper compiler options are known (currently Solaris
1424 and Linux), "threads" is the default.
1427 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
1430 *) Install various scripts to $(OPENSSLDIR)/misc, not to
1431 $(INSTALLTOP)/bin -- they shouldn't clutter directories
1432 such as /usr/local/bin.
1435 *) "make linux-shared" to build shared libraries.
1436 [Niels Poppe <niels@netbox.org>]
1438 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
1441 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
1442 extension adding in x509 utility.
1445 *) Remove NOPROTO sections and error code comments.
1448 *) Partial rewrite of the DEF file generator to now parse the ANSI
1452 *) New Configure options --prefix=DIR and --openssldir=DIR.
1455 *) Complete rewrite of the error code script(s). It is all now handled
1456 by one script at the top level which handles error code gathering,
1457 header rewriting and C source file generation. It should be much better
1458 than the old method: it now uses a modified version of Ulf's parser to
1459 read the ANSI prototypes in all header files (thus the old K&R definitions
1460 aren't needed for error creation any more) and do a better job of
1461 translating function codes into names. The old 'ASN1 error code imbedded
1462 in a comment' is no longer necessary and it doesn't use .err files which
1463 have now been deleted. Also the error code call doesn't have to appear all
1464 on one line (which resulted in some large lines...).
1467 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
1470 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
1471 0 (which usually indicates a closed connection), but continue reading.
1474 *) Fix some race conditions.
1477 *) Add support for CRL distribution points extension. Add Certificate
1478 Policies and CRL distribution points documentation.
1481 *) Move the autogenerated header file parts to crypto/opensslconf.h.
1484 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
1485 8 of keying material. Merlin has also confirmed interop with this fix
1486 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
1487 [Merlin Hughes <merlin@baltimore.ie>]
1489 *) Fix lots of warnings.
1490 [Richard Levitte <levitte@stacken.kth.se>]
1492 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
1493 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
1494 [Richard Levitte <levitte@stacken.kth.se>]
1496 *) Fix problems with sizeof(long) == 8.
1497 [Andy Polyakov <appro@fy.chalmers.se>]
1499 *) Change functions to ANSI C.
1502 *) Fix typos in error codes.
1503 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
1505 *) Remove defunct assembler files from Configure.
1508 *) SPARC v8 assembler BIGNUM implementation.
1509 [Andy Polyakov <appro@fy.chalmers.se>]
1511 *) Support for Certificate Policies extension: both print and set.
1512 Various additions to support the r2i method this uses.
1515 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
1516 return a const string when you are expecting an allocated buffer.
1519 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
1520 types DirectoryString and DisplayText.
1523 *) Add code to allow r2i extensions to access the configuration database,
1524 add an LHASH database driver and add several ctx helper functions.
1527 *) Fix an evil bug in bn_expand2() which caused various BN functions to
1528 fail when they extended the size of a BIGNUM.
1531 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
1532 support typesafe stack.
1535 *) Fix typo in SSL_[gs]et_options().
1536 [Nils Frostberg <nils@medcom.se>]
1538 *) Delete various functions and files that belonged to the (now obsolete)
1539 old X509V3 handling code.
1542 *) New Configure option "rsaref".
1545 *) Don't auto-generate pem.h.
1548 *) Introduce type-safe ASN.1 SETs.
1551 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
1552 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
1554 *) Introduce type-safe STACKs. This will almost certainly break lots of code
1555 that links with OpenSSL (well at least cause lots of warnings), but fear
1556 not: the conversion is trivial, and it eliminates loads of evil casts. A
1557 few STACKed things have been converted already. Feel free to convert more.
1558 In the fullness of time, I'll do away with the STACK type altogether.
1561 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
1562 specified in <certfile> by updating the entry in the index.txt file.
1563 This way one no longer has to edit the index.txt file manually for
1564 revoking a certificate. The -revoke option does the gory details now.
1565 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
1567 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
1568 `-text' option at all and this way the `-noout -text' combination was
1569 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
1570 [Ralf S. Engelschall]
1572 *) Make sure a corresponding plain text error message exists for the
1573 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
1574 verify callback function determined that a certificate was revoked.
1575 [Ralf S. Engelschall]
1577 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
1578 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
1579 all available cipers including rc5, which was forgotten until now.
1580 In order to let the testing shell script know which algorithms
1581 are available, a new (up to now undocumented) command
1582 "openssl list-cipher-commands" is used.
1585 *) Bugfix: s_client occasionally would sleep in select() when
1586 it should have checked SSL_pending() first.
1589 *) New functions DSA_do_sign and DSA_do_verify to provide access to
1590 the raw DSA values prior to ASN.1 encoding.
1593 *) Tweaks to Configure
1594 [Niels Poppe <niels@netbox.org>]
1596 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
1600 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
1603 *) New config option to avoid instructions that are illegal on the 80386.
1604 The default code is faster, but requires at least a 486.
1607 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
1608 SSL2_SERVER_VERSION (not used at all) macros, which are now the
1609 same as SSL2_VERSION anyway.
1612 *) New "-showcerts" option for s_client.
1615 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
1616 application. Various cleanups and fixes.
1619 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
1620 modify error routines to work internally. Add error codes and PBE init
1621 to library startup routines.
1624 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
1625 packing functions to asn1 and evp. Changed function names and error
1626 codes along the way.
1629 *) PKCS12 integration: and so it begins... First of several patches to
1630 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
1631 objects to objects.h
1634 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
1635 and display support for Thawte strong extranet extension.
1638 *) Add LinuxPPC support.
1639 [Jeff Dubrule <igor@pobox.org>]
1641 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
1642 bn_div_words in alpha.s.
1643 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
1645 *) Make sure the RSA OAEP test is skipped under -DRSAref because
1646 OAEP isn't supported when OpenSSL is built with RSAref.
1647 [Ulf Moeller <ulf@fitug.de>]
1649 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
1650 so they no longer are missing under -DNOPROTO.
1651 [Soren S. Jorvang <soren@t.dk>]
1654 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
1656 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
1657 doesn't work when the session is reused. Coming soon!
1660 *) Fix a security hole, that allows sessions to be reused in the wrong
1661 context thus bypassing client cert protection! All software that uses
1662 client certs and session caches in multiple contexts NEEDS PATCHING to
1663 allow session reuse! A fuller solution is in the works.
1664 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
1666 *) Some more source tree cleanups (removed obsolete files
1667 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
1668 permission on "config" script to be executable) and a fix for the INSTALL
1670 [Ulf Moeller <ulf@fitug.de>]
1672 *) Remove some legacy and erroneous uses of malloc, free instead of
1674 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
1676 *) Make rsa_oaep_test return non-zero on error.
1677 [Ulf Moeller <ulf@fitug.de>]
1679 *) Add support for native Solaris shared libraries. Configure
1680 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
1681 if someone would make that last step automatic.
1682 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
1684 *) ctx_size was not built with the right compiler during "make links". Fixed.
1687 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
1688 except NULL ciphers". This means the default cipher list will no longer
1689 enable NULL ciphers. They need to be specifically enabled e.g. with
1690 the string "DEFAULT:eNULL".
1693 *) Fix to RSA private encryption routines: if p < q then it would
1694 occasionally produce an invalid result. This will only happen with
1695 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
1698 *) Be less restrictive and allow also `perl util/perlpath.pl
1699 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
1700 because this way one can also use an interpreter named `perl5' (which is
1701 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
1702 installed as `perl').
1703 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
1705 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
1706 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
1708 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
1709 advapi32.lib to Win32 build and change the pem test comparision
1710 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
1711 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
1712 and crypto/des/ede_cbcm_enc.c.
1715 *) DES quad checksum was broken on big-endian architectures. Fixed.
1718 *) Comment out two functions in bio.h that aren't implemented. Fix up the
1719 Win32 test batch file so it (might) work again. The Win32 test batch file
1720 is horrible: I feel ill....
1723 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
1724 in e_os.h. Audit of header files to check ANSI and non ANSI
1725 sections: 10 functions were absent from non ANSI section and not exported
1726 from Windows DLLs. Fixed up libeay.num for new functions.
1729 *) Make `openssl version' output lines consistent.
1730 [Ralf S. Engelschall]
1732 *) Fix Win32 symbol export lists for BIO functions: Added
1733 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
1734 to ms/libeay{16,32}.def.
1735 [Ralf S. Engelschall]
1737 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
1738 fine under Unix and passes some trivial tests I've now added. But the
1739 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
1740 added to make sure no one expects that this stuff really works in the
1741 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
1742 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
1744 [Ralf S. Engelschall]
1746 *) Fix the generation of two part addresses in perl.
1747 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
1749 *) Add config entry for Linux on MIPS.
1750 [John Tobey <jtobey@channel1.com>]
1752 *) Make links whenever Configure is run, unless we are on Windoze.
1755 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
1756 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
1760 *) Add a useful kludge to allow package maintainers to specify compiler and
1761 other platforms details on the command line without having to patch the
1762 Configure script everytime: One now can use ``perl Configure
1763 <id>:<details>'', i.e. platform ids are allowed to have details appended
1764 to them (seperated by colons). This is treated as there would be a static
1765 pre-configured entry in Configure's %table under key <id> with value
1766 <details> and ``perl Configure <id>'' is called. So, when you want to
1767 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
1768 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
1769 now, which overrides the FreeBSD-elf entry on-the-fly.
1770 [Ralf S. Engelschall]
1772 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
1775 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
1776 on the `perl Configure ...' command line. This way one can compile
1777 OpenSSL libraries with Position Independent Code (PIC) which is needed
1778 for linking it into DSOs.
1779 [Ralf S. Engelschall]
1781 *) Remarkably, export ciphers were totally broken and no-one had noticed!
1785 *) Cleaned up the LICENSE document: The official contact for any license
1786 questions now is the OpenSSL core team under openssl-core@openssl.org.
1787 And add a paragraph about the dual-license situation to make sure people
1788 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
1789 to the OpenSSL toolkit.
1790 [Ralf S. Engelschall]
1792 *) General source tree makefile cleanups: Made `making xxx in yyy...'
1793 display consistent in the source tree and replaced `/bin/rm' by `rm'.
1794 Additonally cleaned up the `make links' target: Remove unnecessary
1795 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
1796 to speed processing and no longer clutter the display with confusing
1797 stuff. Instead only the actually done links are displayed.
1798 [Ralf S. Engelschall]
1800 *) Permit null encryption ciphersuites, used for authentication only. It used
1801 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
1802 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
1806 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
1807 signed attributes when verifying signatures (this would break them),
1808 the detached data encoding was wrong and public keys obtained using
1809 X509_get_pubkey() weren't freed.
1812 *) Add text documentation for the BUFFER functions. Also added a work around
1813 to a Win95 console bug. This was triggered by the password read stuff: the
1814 last character typed gets carried over to the next fread(). If you were
1815 generating a new cert request using 'req' for example then the last
1816 character of the passphrase would be CR which would then enter the first
1820 *) Added the new `Includes OpenSSL Cryptography Software' button as
1821 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
1822 button and can be used by applications based on OpenSSL to show the
1823 relationship to the OpenSSL project.
1824 [Ralf S. Engelschall]
1826 *) Remove confusing variables in function signatures in files
1827 ssl/ssl_lib.c and ssl/ssl.h.
1828 [Lennart Bong <lob@kulthea.stacken.kth.se>]
1830 *) Don't install bss_file.c under PREFIX/include/
1831 [Lennart Bong <lob@kulthea.stacken.kth.se>]
1833 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
1834 functions that return function pointers and has support for NT specific
1835 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
1836 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
1837 unsigned to signed types: this was killing the Win32 compile.
1840 *) Add new certificate file to stack functions,
1841 SSL_add_dir_cert_subjects_to_stack() and
1842 SSL_add_file_cert_subjects_to_stack(). These largely supplant
1843 SSL_load_client_CA_file(), and can be used to add multiple certs easily
1844 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
1845 This means that Apache-SSL and similar packages don't have to mess around
1846 to add as many CAs as they want to the preferred list.
1849 *) Experiment with doxygen documentation. Currently only partially applied to
1851 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
1852 openssl.doxy as the configuration file.
1855 *) Get rid of remaining C++-style comments which strict C compilers hate.
1856 [Ralf S. Engelschall, pointed out by Carlos Amengual]
1858 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
1859 compiled in by default: it has problems with large keys.
1862 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
1863 DH private keys and/or callback functions which directly correspond to
1864 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
1865 is needed for applications which have to configure certificates on a
1866 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
1868 For the RSA certificate situation is makes no difference, but
1869 for the DSA certificate situation this fixes the "no shared cipher"
1870 problem where the OpenSSL cipher selection procedure failed because the
1871 temporary keys were not overtaken from the context and the API provided
1872 no way to reconfigure them.
1873 The new functions now let applications reconfigure the stuff and they
1874 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
1875 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
1876 non-public-API function ssl_cert_instantiate() is used as a helper
1877 function and also to reduce code redundancy inside ssl_rsa.c.
1878 [Ralf S. Engelschall]
1880 *) Move s_server -dcert and -dkey options out of the undocumented feature
1881 area because they are useful for the DSA situation and should be
1882 recognized by the users.
1883 [Ralf S. Engelschall]
1885 *) Fix the cipher decision scheme for export ciphers: the export bits are
1886 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
1887 SSL_EXP_MASK. So, the original variable has to be used instead of the
1888 already masked variable.
1889 [Richard Levitte <levitte@stacken.kth.se>]
1891 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
1892 [Richard Levitte <levitte@stacken.kth.se>]
1894 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
1895 from `int' to `unsigned int' because it's a length and initialized by
1896 EVP_DigestFinal() which expects an `unsigned int *'.
1897 [Richard Levitte <levitte@stacken.kth.se>]
1899 *) Don't hard-code path to Perl interpreter on shebang line of Configure
1900 script. Instead use the usual Shell->Perl transition trick.
1901 [Ralf S. Engelschall]
1903 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
1904 (in addition to RSA certificates) to match the behaviour of `openssl dsa
1905 -noout -modulus' as it's already the case for `openssl rsa -noout
1906 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
1907 currently the public key is printed (a decision which was already done by
1908 `openssl dsa -modulus' in the past) which serves a similar purpose.
1909 Additionally the NO_RSA no longer completely removes the whole -modulus
1910 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
1912 [Ralf S. Engelschall]
1914 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
1915 BIO. See the source (crypto/evp/bio_ok.c) for more info.
1916 [Arne Ansper <arne@ats.cyber.ee>]
1918 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
1919 to be added. Now both 'req' and 'ca' can use new objects defined in the
1923 *) Add cool BIO that does syslog (or event log on NT).
1924 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
1926 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
1927 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
1928 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
1929 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
1932 *) Add preliminary config info for new extension code.
1935 *) Make RSA_NO_PADDING really use no padding.
1936 [Ulf Moeller <ulf@fitug.de>]
1938 *) Generate errors when private/public key check is done.
1941 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
1942 for some CRL extensions and new objects added.
1945 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
1946 key usage extension and fuller support for authority key id.
1949 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
1950 padding method for RSA, which is recommended for new applications in PKCS
1951 #1 v2.0 (RFC 2437, October 1998).
1952 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
1953 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
1954 against Bleichbacher's attack on RSA.
1955 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
1958 *) Updates to the new SSL compression code
1959 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
1961 *) Fix so that the version number in the master secret, when passed
1962 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
1963 (because the server will not accept higher), that the version number
1964 is 0x03,0x01, not 0x03,0x00
1965 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
1967 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
1968 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
1969 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
1972 *) Support for RAW extensions where an arbitrary extension can be
1973 created by including its DER encoding. See apps/openssl.cnf for
1977 *) Make sure latest Perl versions don't interpret some generated C array
1978 code as Perl array code in the crypto/err/err_genc.pl script.
1979 [Lars Weber <3weber@informatik.uni-hamburg.de>]
1981 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
1982 not many people have the assembler. Various Win32 compilation fixes and
1983 update to the INSTALL.W32 file with (hopefully) more accurate Win32
1987 *) Modify configure script 'Configure' to automatically create crypto/date.h
1988 file under Win32 and also build pem.h from pem.org. New script
1989 util/mkfiles.pl to create the MINFO file on environments that can't do a
1990 'make files': perl util/mkfiles.pl >MINFO should work.
1993 *) Major rework of DES function declarations, in the pursuit of correctness
1994 and purity. As a result, many evil casts evaporated, and some weirdness,
1995 too. You may find this causes warnings in your code. Zapping your evil
1996 casts will probably fix them. Mostly.
1999 *) Fix for a typo in asn1.h. Bug fix to object creation script
2000 obj_dat.pl. It considered a zero in an object definition to mean
2001 "end of object": none of the objects in objects.h have any zeros
2002 so it wasn't spotted.
2003 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2005 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2006 Masking (CBCM). In the absence of test vectors, the best I have been able
2007 to do is check that the decrypt undoes the encrypt, so far. Send me test
2008 vectors if you have them.
2011 *) Correct calculation of key length for export ciphers (too much space was
2012 allocated for null ciphers). This has not been tested!
2015 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2016 message is now correct (it understands "crypto" and "ssl" on its
2017 command line). There is also now an "update" option. This will update
2018 the util/ssleay.num and util/libeay.num files with any new functions.
2020 perl util/mkdef.pl crypto ssl update
2021 it will update them.
2024 *) Overhauled the Perl interface (perl/*):
2025 - ported BN stuff to OpenSSL's different BN library
2026 - made the perl/ source tree CVS-aware
2027 - renamed the package from SSLeay to OpenSSL (the files still contain
2028 their history because I've copied them in the repository)
2029 - removed obsolete files (the test scripts will be replaced
2030 by better Test::Harness variants in the future)
2031 [Ralf S. Engelschall]
2033 *) First cut for a very conservative source tree cleanup:
2034 1. merge various obsolete readme texts into doc/ssleay.txt
2035 where we collect the old documents and readme texts.
2036 2. remove the first part of files where I'm already sure that we no
2037 longer need them because of three reasons: either they are just temporary
2038 files which were left by Eric or they are preserved original files where
2039 I've verified that the diff is also available in the CVS via "cvs diff
2040 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2041 the crypto/md/ stuff).
2042 [Ralf S. Engelschall]
2044 *) More extension code. Incomplete support for subject and issuer alt
2045 name, issuer and authority key id. Change the i2v function parameters
2046 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2047 what that's for :-) Fix to ASN1 macro which messed up
2048 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2051 *) Preliminary support for ENUMERATED type. This is largely copied from the
2055 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2056 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2058 *) Make sure `make rehash' target really finds the `openssl' program.
2059 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2061 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2062 like to hear about it if this slows down other processors.
2065 *) Add CygWin32 platform information to Configure script.
2066 [Alan Batie <batie@aahz.jf.intel.com>]
2068 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2069 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2071 *) New program nseq to manipulate netscape certificate sequences
2074 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2078 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2079 but the BN code had some problems that would cause failures when
2080 doing certificate verification and some other functions.
2081 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2083 *) Add ASN1 and PEM code to support netscape certificate sequences.
2086 *) Add ASN1 and PEM code to support netscape certificate sequences.
2089 *) Add several PKIX and private extended key usage OIDs.
2092 *) Modify the 'ca' program to handle the new extension code. Modify
2093 openssl.cnf for new extension format, add comments.
2096 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2097 and add a sample to openssl.cnf so req -x509 now adds appropriate
2101 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2102 error code, add initial support to X509_print() and x509 application.
2105 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2106 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2107 stuff is currently isolated and isn't even compiled yet.
2110 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2111 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2112 Removed the versions check from X509 routines when loading extensions:
2113 this allows certain broken certificates that don't set the version
2114 properly to be processed.
2117 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2118 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2119 can still be regenerated with "make depend".
2122 *) Spelling mistake in C version of CAST-128.
2123 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2125 *) Changes to the error generation code. The perl script err-code.pl
2126 now reads in the old error codes and retains the old numbers, only
2127 adding new ones if necessary. It also only changes the .err files if new
2128 codes are added. The makefiles have been modified to only insert errors
2129 when needed (to avoid needlessly modifying header files). This is done
2130 by only inserting errors if the .err file is newer than the auto generated
2131 C file. To rebuild all the error codes from scratch (the old behaviour)
2132 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2133 or delete all the .err files.
2136 *) CAST-128 was incorrectly implemented for short keys. The C version has
2137 been fixed, but is untested. The assembler versions are also fixed, but
2138 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2139 to regenerate it if needed.
2140 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2141 Hagino <itojun@kame.net>]
2143 *) File was opened incorrectly in randfile.c.
2144 [Ulf Möller <ulf@fitug.de>]
2146 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2147 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2148 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2149 al: it's just almost always a UTCTime. Note this patch adds new error
2150 codes so do a "make errors" if there are problems.
2153 *) Correct Linux 1 recognition in config.
2154 [Ulf Möller <ulf@fitug.de>]
2156 *) Remove pointless MD5 hash when using DSA keys in ca.
2157 [Anonymous <nobody@replay.com>]
2159 *) Generate an error if given an empty string as a cert directory. Also
2160 generate an error if handed NULL (previously returned 0 to indicate an
2161 error, but didn't set one).
2162 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
2164 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
2167 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
2168 parameters. This was causing a warning which killed off the Win32 compile.
2171 *) Remove C++ style comments from crypto/bn/bn_local.h.
2172 [Neil Costigan <neil.costigan@celocom.com>]
2174 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
2175 based on a text string, looking up short and long names and finally
2176 "dot" format. The "dot" format stuff didn't work. Added new function
2177 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
2178 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
2179 OID is not part of the table.
2182 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
2183 X509_LOOKUP_by_alias().
2186 *) Sort openssl functions by name.
2189 *) Get the gendsa program working (hopefully) and add it to app list. Remove
2190 encryption from sample DSA keys (in case anyone is interested the password
2194 *) Make _all_ *_free functions accept a NULL pointer.
2195 [Frans Heymans <fheymans@isaserver.be>]
2197 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
2199 [Anonymous <nobody@replay.com>]
2201 *) s_server should send the CAfile as acceptable CAs, not its own cert.
2202 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2204 *) Don't blow it for numeric -newkey arguments to apps/req.
2205 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2207 *) Temp key "for export" tests were wrong in s3_srvr.c.
2208 [Anonymous <nobody@replay.com>]
2210 *) Add prototype for temp key callback functions
2211 SSL_CTX_set_tmp_{rsa,dh}_callback().
2214 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
2215 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
2218 *) X509_name_add_entry() freed the wrong thing after an error.
2219 [Arne Ansper <arne@ats.cyber.ee>]
2221 *) rsa_eay.c would attempt to free a NULL context.
2222 [Arne Ansper <arne@ats.cyber.ee>]
2224 *) BIO_s_socket() had a broken should_retry() on Windoze.
2225 [Arne Ansper <arne@ats.cyber.ee>]
2227 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
2228 [Arne Ansper <arne@ats.cyber.ee>]
2230 *) Make sure the already existing X509_STORE->depth variable is initialized
2231 in X509_STORE_new(), but document the fact that this variable is still
2232 unused in the certificate verification process.
2233 [Ralf S. Engelschall]
2235 *) Fix the various library and apps files to free up pkeys obtained from
2236 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
2239 *) Fix reference counting in X509_PUBKEY_get(). This makes
2240 demos/maurice/example2.c work, amongst others, probably.
2241 [Steve Henson and Ben Laurie]
2243 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
2244 `openssl' and second, the shortcut symlinks for the `openssl <command>'
2245 are no longer created. This way we have a single and consistent command
2246 line interface `openssl <command>', similar to `cvs <command>'.
2247 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
2249 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
2250 BIT STRING wrapper always have zero unused bits.
2253 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
2256 *) Make the top-level INSTALL documentation easier to understand.
2259 *) Makefiles updated to exit if an error occurs in a sub-directory
2260 make (including if user presses ^C) [Paul Sutton]
2262 *) Make Montgomery context stuff explicit in RSA data structure.
2265 *) Fix build order of pem and err to allow for generated pem.h.
2268 *) Fix renumbering bug in X509_NAME_delete_entry().
2271 *) Enhanced the err-ins.pl script so it makes the error library number
2272 global and can add a library name. This is needed for external ASN1 and
2273 other error libraries.
2276 *) Fixed sk_insert which never worked properly.
2279 *) Fix ASN1 macros so they can handle indefinite length construted
2280 EXPLICIT tags. Some non standard certificates use these: they can now
2284 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
2285 into a single doc/ssleay.txt bundle. This way the information is still
2286 preserved but no longer messes up this directory. Now it's new room for
2287 the new set of documenation files.
2288 [Ralf S. Engelschall]
2290 *) SETs were incorrectly DER encoded. This was a major pain, because they
2291 shared code with SEQUENCEs, which aren't coded the same. This means that
2292 almost everything to do with SETs or SEQUENCEs has either changed name or
2293 number of arguments.
2294 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
2296 *) Fix test data to work with the above.
2299 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
2300 was already fixed by Eric for 0.9.1 it seems.
2301 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
2303 *) Autodetect FreeBSD3.
2306 *) Fix various bugs in Configure. This affects the following platforms:
2310 unixware-2.0-pentium
2314 *) Eliminate generated files from CVS. Reorder tests to regenerate files
2315 before they are needed.
2318 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
2322 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
2324 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
2325 changed SSLeay to OpenSSL in version strings.
2326 [Ralf S. Engelschall]
2328 *) Some fixups to the top-level documents.
2331 *) Fixed the nasty bug where rsaref.h was not found under compile-time
2332 because the symlink to include/ was missing.
2333 [Ralf S. Engelschall]
2335 *) Incorporated the popular no-RSA/DSA-only patches
2336 which allow to compile a RSA-free SSLeay.
2337 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
2339 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
2340 when "ssleay" is still not found.
2341 [Ralf S. Engelschall]
2343 *) Added more platforms to Configure: Cray T3E, HPUX 11,
2344 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
2346 *) Updated the README file.
2347 [Ralf S. Engelschall]
2349 *) Added various .cvsignore files in the CVS repository subdirs
2350 to make a "cvs update" really silent.
2351 [Ralf S. Engelschall]
2353 *) Recompiled the error-definition header files and added
2354 missing symbols to the Win32 linker tables.
2355 [Ralf S. Engelschall]
2357 *) Cleaned up the top-level documents;
2358 o new files: CHANGES and LICENSE
2359 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
2360 o merged COPYRIGHT into LICENSE
2361 o removed obsolete TODO file
2362 o renamed MICROSOFT to INSTALL.W32
2363 [Ralf S. Engelschall]
2365 *) Removed dummy files from the 0.9.1b source tree:
2366 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
2367 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
2368 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
2369 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
2370 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
2371 [Ralf S. Engelschall]
2373 *) Added various platform portability fixes.
2376 *) The Genesis of the OpenSSL rpject:
2377 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
2378 Young and Tim J. Hudson created while they were working for C2Net until
2380 [The OpenSSL Project]
2383 Changes between 0.9.0b and 0.9.1b [not released]
2385 *) Updated a few CA certificates under certs/
2388 *) Changed some BIGNUM api stuff.
2391 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
2392 DGUX x86, Linux Alpha, etc.
2395 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
2396 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
2400 *) Add -strparse option to asn1pars program which parses nested
2402 [Dr Stephen Henson <shenson@bigfoot.com>]
2404 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
2407 *) DSA fix for "ca" program.
2410 *) Added "-genkey" option to "dsaparam" program.
2413 *) Added RIPE MD160 (rmd160) message digest.
2416 *) Added -a (all) option to "ssleay version" command.
2419 *) Added PLATFORM define which is the id given to Configure.
2422 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
2425 *) Extended the ASN.1 parser routines.
2428 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
2431 *) Added a BN_CTX to the BN library.
2434 *) Fixed the weak key values in DES library
2437 *) Changed API in EVP library for cipher aliases.
2440 *) Added support for RC2/64bit cipher.
2443 *) Converted the lhash library to the crypto/mem.c functions.
2446 *) Added more recognized ASN.1 object ids.
2449 *) Added more RSA padding checks for SSL/TLS.
2452 *) Added BIO proxy/filter functionality.
2455 *) Added extra_certs to SSL_CTX which can be used
2456 send extra CA certificates to the client in the CA cert chain sending
2457 process. It can be configured with SSL_CTX_add_extra_chain_cert().
2460 *) Now Fortezza is denied in the authentication phase because
2461 this is key exchange mechanism is not supported by SSLeay at all.
2464 *) Additional PKCS1 checks.
2467 *) Support the string "TLSv1" for all TLS v1 ciphers.
2470 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
2471 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
2474 *) Fixed a few memory leaks.
2477 *) Fixed various code and comment typos.
2480 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
2481 bytes sent in the client random.
2482 [Edward Bishop <ebishop@spyglass.com>]