5 Changes between 0.9.4 and 0.9.5 [xx XXX 1999]
7 *) Rebuild of the memory allocation routines used by OpenSSL code and
8 possibly others as well. The purpose is to make an interface that
9 provide hooks so anyone can build a separate set of allocation and
10 deallocation routines to be used by OpenSSL, for example if memory
11 pool implementations, or something else. The same is provided for
12 memory debugging code. OpenSSL already comes with code that finds
13 memory leaks, but this gives people a chance to debug other memory
17 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
18 ordering of SMIMECapabilities wasn't in "strength order" and there
19 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
23 *) Some ASN1 types with illegal zero length encoding (INTEGER,
24 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
25 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
27 *) Merge in my S/MIME library for OpenSSL. This provides a simple
28 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
29 functionality to handle multipart/signed properly) and a utility
30 called 'smime' to call all this stuff. This is based on code I
31 originally wrote for Celo who have kindly allowed it to be
35 *) Add variants des_set_key_checked and des_set_key_unchecked of
36 des_set_key (aka des_key_sched). Global variable des_check_key
37 decides which of these is called by des_set_key; this way
38 des_check_key behaves as it always did, but applications and
39 the library itself, which was buggy for des_check_key == 1,
40 have a cleaner way to pick the version they need.
43 *) New function PKCS12_newpass() which changes the password of a
47 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
48 dynamic mix. In both cases the ids can be used as an index into the
49 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
50 functions so they accept a list of the field values and the
51 application doesn't need to directly manipulate the X509_TRUST
55 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
59 *) Modify the way the V3 extension code looks up extensions. This now
60 works in a similar way to the object code: we have some "standard"
61 extensions in a static table which is searched with OBJ_bsearch()
62 and the application can add dynamic ones if needed. The file
63 crypto/x509v3/ext_dat.h now has the info: this file needs to be
64 updated whenever a new extension is added to the core code and kept
65 in ext_nid order. There is a simple program 'tabtest.c' which checks
66 this. New extensions are not added too often so this file can readily
67 be maintained manually.
69 There are two big advantages in doing things this way. The extensions
70 can be looked up immediately and no longer need to be "added" using
71 X509V3_add_standard_extensions(): this function now does nothing.
72 [Side note: I get *lots* of email saying the extension code doesn't
73 work because people forget to call this function]
74 Also no dynamic allocation is done unless new extensions are added:
75 so if we don't add custom extensions there is no need to call
79 *) Modify enc utility's salting as follows: make salting the default. Add a
80 magic header, so unsalted files fail gracefully instead of just decrypting
81 to garbage. This is because not salting is a big security hole, so people
82 should be discouraged from doing it.
85 *) Fixes and enhancements to the 'x509' utility. It allowed a message
86 digest to be passed on the command line but it only used this
87 parameter when signing a certificate. Modified so all relevant
88 operations are affected by the digest parameter including the
89 -fingerprint and -x509toreq options. Also -x509toreq choked if a
90 DSA key was used because it didn't fix the digest.
93 *) Initial certificate chain verify code. Currently tests the untrusted
94 certificates for consistency with the verify purpose (which is set
95 when the X509_STORE_CTX structure is set up) and checks the pathlength.
97 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
98 this is because it will reject chains with invalid extensions whereas
99 every previous version of OpenSSL and SSLeay made no checks at all.
101 Trust code: checks the root CA for the relevant trust settings. Trust
102 settings have an initial value consistent with the verify purpose: e.g.
103 if the verify purpose is for SSL client use it expects the CA to be
104 trusted for SSL client use. However the default value can be changed to
105 permit custom trust settings: one example of this would be to only trust
106 certificates from a specific "secure" set of CAs.
108 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
109 which should be used for version portability: especially since the
110 verify structure is likely to change more often now.
112 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
113 to set them. If not set then assume SSL clients will verify SSL servers
116 Two new options to the verify program: -untrusted allows a set of
117 untrusted certificates to be passed in and -purpose which sets the
118 intended purpose of the certificate. If a purpose is set then the
119 new chain verify code is used to check extension consistency.
122 *) Support for the authority information access extension.
125 *) Modify RSA and DSA PEM read routines to transparently handle
126 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
127 public keys in a format compatible with certificate
128 SubjectPublicKeyInfo structures. Unfortunately there were already
129 functions called *_PublicKey_* which used various odd formats so
130 these are retained for compatability: however the DSA variants were
131 never in a public release so they have been deleted. Changed dsa/rsa
132 utilities to handle the new format: note no releases ever handled public
133 keys so we should be OK.
135 The primary motivation for this change is to avoid the same fiasco
136 that dogs private keys: there are several incompatible private key
137 formats some of which are standard and some OpenSSL specific and
138 require various evil hacks to allow partial transparent handling and
139 even then it doesn't work with DER formats. Given the option anything
140 other than PKCS#8 should be dumped: but the other formats have to
141 stay in the name of compatability.
143 With public keys and the benefit of hindsight one standard format
144 is used which works with EVP_PKEY, RSA or DSA structures: though
145 it clearly returns an error if you try to read the wrong kind of key.
147 Added a -pubkey option to the 'x509' utility to output the public key.
148 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*() and add
149 EVP_PKEY_rset_*() functions that do the same as the EVP_PKEY_assign_*()
150 except they up the reference count of the added key (they don't "swallow"
154 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
155 CRLs would fail if the file contained no certificates or no CRLs:
156 added a new function to read in both types and return the number
157 read: this means that if none are read it will be an error. The
158 DER versions of the certificate and CRL reader would always fail
159 because it isn't possible to mix certificates and CRLs in DER format
160 without choking one or the other routine. Changed this to just read
161 a certificate: this is the best we can do. Also modified the code
162 in apps/verify.c to take notice of return codes: it was previously
163 attempting to read in certificates from NULL pointers and ignoring
164 any errors: this is one reason why the cert and CRL reader seemed
165 to work. It doesn't check return codes from the default certificate
166 routines: these may well fail if the certificates aren't installed.
169 *) Code to support otherName option in GeneralName.
172 *) First update to verify code. Change the verify utility
173 so it warns if it is passed a self signed certificate:
174 for consistency with the normal behaviour. X509_verify
175 has been modified to it will now verify a self signed
176 certificate if *exactly* the same certificate appears
177 in the store: it was previously impossible to trust a
178 single self signed certificate. This means that:
179 openssl verify ss.pem
180 now gives a warning about a self signed certificate but
181 openssl verify -CAfile ss.pem ss.pem
185 *) For servers, store verify_result in SSL_SESSION data structure
186 (and add it to external session representation).
187 This is needed when client certificate verifications fails,
188 but an application-provided verification callback (set by
189 SSL_CTX_set_cert_verify_callback) allows accepting the session
190 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
191 but returns 1): When the session is reused, we have to set
192 ssl->verify_result to the appropriate error code to avoid
194 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
196 *) Fix a bug in the new PKCS#7 code: it didn't consider the
197 case in PKCS7_dataInit() where the signed PKCS7 structure
198 didn't contain any existing data because it was being created.
199 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
201 *) Add a salt to the key derivation routines in enc.c. This
202 forms the first 8 bytes of the encrypted file. Also add a
203 -S option to allow a salt to be input on the command line.
206 *) New function X509_cmp(). Oddly enough there wasn't a function
207 to compare two certificates. We do this by working out the SHA1
208 hash and comparing that. X509_cmp() will be needed by the trust
212 *) Correctly increment the reference count in the SSL_SESSION pointer
213 returned from SSL_get_session().
214 [Geoff Thorpe <geoff@eu.c2.net>]
216 *) Fix for 'req': it was adding a null to request attributes.
217 Also change the X509_LOOKUP and X509_INFO code to handle
218 certificate auxiliary information.
221 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
225 *) Add the possibility to add extra information to the memory leak
226 detecting output, to form tracebacks, showing from where each
227 allocation was originated. Also updated sid code to be multi-
231 *) Add options -text and -noout to pkcs7 utility and delete the
232 encryption options which never did anything. Update docs.
235 *) Add options to some of the utilities to allow the pass phrase
236 to be included on either the command line (not recommended on
237 OSes like Unix) or read from the environment. Update the
238 manpages and fix a few bugs.
241 *) Add a few manpages for some of the openssl commands.
244 *) Fix the -revoke option in ca. It was freeing up memory twice,
245 leaking and not finding already revoked certificates.
248 *) Extensive changes to support certificate auxiliary information.
249 This involves the use of X509_CERT_AUX structure and X509_AUX
250 functions. An X509_AUX function such as PEM_read_X509_AUX()
251 can still read in a certificate file in the usual way but it
252 will also read in any additional "auxiliary information". By
253 doing things this way a fair degree of compatability can be
254 retained: existing certificates can have this information added
255 using the new 'x509' options.
257 Current auxiliary information includes an "alias" and some trust
258 settings. The trust settings will ultimately be used in enhanced
259 certificate chain verification routines: currently a certificate
260 can only be trusted if it is self signed and then it is trusted
264 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD). The
265 problem was that one of the replacement routines had not been working since
266 SSLeay releases. For now the offending routine has been replaced with
267 non-optimised assembler. Even so, this now gives around 95% performance
268 improvement for 1024 bit RSA signs.
271 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
272 handling. Most clients have the effective key size in bits equal to
273 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
274 A few however don't do this and instead use the size of the decrypted key
275 to determine the RC2 key length and the AlgorithmIdentifier to determine
276 the effective key length. In this case the effective key lenth can still
277 be 40 bits but the key length can be 168 bits for example. This is fixed
278 by manually forcing an RC2 key into the EVP_PKEY structure because the
279 EVP code can't currently handle unusual RC2 key sizes: it always assumes
280 the key length and effective key length are equal.
283 *) Add a bunch of functions that should simplify the creation of
284 X509_NAME structures. Now you should be able to do:
285 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
286 and have it automatically work out the correct field type and fill in
287 the structures. The more adventurous can try:
288 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
289 and it will (hopefully) work out the correct multibyte encoding.
292 *) Change the 'req' utility to use the new field handling and multibyte
293 copy routines. Before the DN field creation was handled in an ad hoc
294 way in req, ca, and x509 which was rather broken and didn't support
295 BMPStrings or UTF8Strings. Since some software doesn't implement
296 BMPStrings or UTF8Strings yet, they can be enabled using the config file
297 using the dirstring_type option. See the new comment in the default
298 openssl.cnf for more info.
301 *) Make crypto/rand/md_rand.c more robust:
302 - Assure unique random numbers after fork().
303 - Make sure that concurrent threads access the global counter and
304 md serializably so that we never lose entropy in them
305 or use exactly the same state in multiple threads.
306 Access to the large state is not always serializable because
307 the additional locking could be a performance killer, and
308 md should be large enough anyway.
311 *) New file apps/app_rand.c with commonly needed functionality
312 for handling the random seed file.
314 Use the random seed file in some applications that previously did not:
316 dsaparam -genkey (which also ignored its `-rand' option),
320 Except on systems with /dev/urandom, it is crucial to have a random
321 seed file at least for key creation, DSA signing, and for DH exchanges;
322 for RSA signatures we could do without one.
324 gendh and gendsa (unlike genrsa) used to read only the first byte
325 of each file listed in the `-rand' option. The function as previously
326 found in genrsa is now in app_rand.c and is used by all programs
327 that support `-rand'.
330 *) In RAND_write_file, use mode 0600 for creating files;
331 don't just chmod when it may be too late.
334 *) Report an error from X509_STORE_load_locations
335 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
338 *) New function ASN1_mbstring_copy() this copies a string in either
339 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
340 into an ASN1_STRING type. A mask of permissible types is passed
341 and it chooses the "minimal" type to use or an error if not type
345 *) Add function equivalents to the various macros in asn1.h. The old
346 macros are retained with an M_ prefix. Code inside the library can
347 use the M_ macros. External code (including the openssl utility)
348 should *NOT* in order to be "shared library friendly".
351 *) Add various functions that can check a certificate's extensions
352 to see if it usable for various purposes such as SSL client,
353 server or S/MIME and CAs of these types. This is currently
354 VERY EXPERIMENTAL but will ultimately be used for certificate chain
355 verification. Also added a -purpose flag to x509 utility to
356 print out all the purposes.
359 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
363 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
364 for, obtain and decode and extension and obtain its critical flag.
365 This allows all the necessary extension code to be handled in a
366 single function call.
369 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
370 platforms. See crypto/rc4/rc4_enc.c for further details.
373 *) New -noout option to asn1parse. This causes no output to be produced
374 its main use is when combined with -strparse and -out to extract data
375 from a file (which may not be in ASN.1 format).
378 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
379 when producing the local key id.
380 [Richard Levitte <levitte@stacken.kth.se>]
382 *) New option -dhparam in s_server. This allows a DH parameter file to be
383 stated explicitly. If it is not stated then it tries the first server
384 certificate file. The previous behaviour hard coded the filename
388 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
389 a public key to be input or output. For example:
390 openssl rsa -in key.pem -pubout -out pubkey.pem
391 Also added necessary DSA public key functions to handle this.
394 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
395 in the message. This was handled by allowing
396 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
397 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
399 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
400 to the end of the strings whereas this didn't. This would cause problems
401 if strings read with d2i_ASN1_bytes() were later modified.
402 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
404 *) Fix for base64 decode bug. When a base64 bio reads only one line of
405 data and it contains EOF it will end up returning an error. This is
406 caused by input 46 bytes long. The cause is due to the way base64
407 BIOs find the start of base64 encoded data. They do this by trying a
408 trial decode on each line until they find one that works. When they
409 do a flag is set and it starts again knowing it can pass all the
410 data directly through the decoder. Unfortunately it doesn't reset
411 the context it uses. This means that if EOF is reached an attempt
412 is made to pass two EOFs through the context and this causes the
413 resulting error. This can also cause other problems as well. As is
414 usual with these problems it takes *ages* to find and the fix is
415 trivial: move one line.
416 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
418 *) Ugly workaround to get s_client and s_server working under Windows. The
419 old code wouldn't work because it needed to select() on sockets and the
420 tty (for keypresses and to see if data could be written). Win32 only
421 supports select() on sockets so we select() with a 1s timeout on the
422 sockets and then see if any characters are waiting to be read, if none
423 are present then we retry, we also assume we can always write data to
424 the tty. This isn't nice because the code then blocks until we've
425 received a complete line of data and it is effectively polling the
426 keyboard at 1s intervals: however it's quite a bit better than not
427 working at all :-) A dedicated Windows application might handle this
428 with an event loop for example.
431 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
432 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
433 will be called when RSA_sign() and RSA_verify() are used. This is useful
434 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
435 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
436 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
437 This necessitated the support of an extra signature type NID_md5_sha1
438 for SSL signatures and modifications to the SSL library to use it instead
439 of calling RSA_public_decrypt() and RSA_private_encrypt().
442 *) Add new -verify -CAfile and -CApath options to the crl program, these
443 will lookup a CRL issuers certificate and verify the signature in a
444 similar way to the verify program. Tidy up the crl program so it
445 no longer acesses structures directly. Make the ASN1 CRL parsing a bit
446 less strict. It will now permit CRL extensions even if it is not
447 a V2 CRL: this will allow it to tolerate some broken CRLs.
450 *) Initialize all non-automatic variables each time one of the openssl
451 sub-programs is started (this is necessary as they may be started
452 multiple times from the "OpenSSL>" prompt).
453 [Lennart Bang, Bodo Moeller]
455 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
456 removing all other RSA functionality (this is what NO_RSA does). This
457 is so (for example) those in the US can disable those operations covered
458 by the RSA patent while allowing storage and parsing of RSA keys and RSA
462 *) Non-copying interface to BIO pairs.
463 (still largely untested)
466 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
467 ASCII string. This was handled independently in various places before.
470 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
471 UTF8 strings a character at a time.
474 *) Use client_version from client hello to select the protocol
475 (s23_srvr.c) and for RSA client key exchange verification
476 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
479 *) Add various utility functions to handle SPKACs, these were previously
480 handled by poking round in the structure internals. Added new function
481 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
482 print, verify and generate SPKACs. Based on an original idea from
483 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
486 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
489 *) Allow the config file extension section to be overwritten on the
490 command line. Based on an original idea from Massimiliano Pala
491 <madwolf@comune.modena.it>. The new option is called -extensions
492 and can be applied to ca, req and x509. Also -reqexts to override
493 the request extensions in req and -crlexts to override the crl extensions
497 *) Add new feature to the SPKAC handling in ca. Now you can include
498 the same field multiple times by preceding it by "XXXX." for example:
501 this is the same syntax as used in the req config file.
504 *) Allow certificate extensions to be added to certificate requests. These
505 are specified in a 'req_extensions' option of the req section of the
506 config file. They can be printed out with the -text option to req but
507 are otherwise ignored at present.
510 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
511 data read consists of only the final block it would not decrypted because
512 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
513 A misplaced 'break' also meant the decrypted final block might not be
514 copied until the next read.
517 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
518 a few extra parameters to the DH structure: these will be useful if
519 for example we want the value of 'q' or implement X9.42 DH.
522 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
523 provides hooks that allow the default DSA functions or functions on a
524 "per key" basis to be replaced. This allows hardware acceleration and
525 hardware key storage to be handled without major modification to the
526 library. Also added low level modexp hooks and CRYPTO_EX structure and
527 associated functions.
530 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
531 as "read only": it can't be written to and the buffer it points to will
532 not be freed. Reading from a read only BIO is much more efficient than
533 a normal memory BIO. This was added because there are several times when
534 an area of memory needs to be read from a BIO. The previous method was
535 to create a memory BIO and write the data to it, this results in two
536 copies of the data and an O(n^2) reading algorithm. There is a new
537 function BIO_new_mem_buf() which creates a read only memory BIO from
538 an area of memory. Also modified the PKCS#7 routines to use read only
542 *) Bugfix: ssl23_get_client_hello did not work properly when called in
543 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
544 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
545 but a retry condition occured while trying to read the rest.
548 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
549 NID_pkcs7_encrypted by default: this was wrong since this should almost
550 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
551 the encrypted data type: this is a more sensible place to put it and it
552 allows the PKCS#12 code to be tidied up that duplicated this
556 *) Changed obj_dat.pl script so it takes its input and output files on
557 the command line. This should avoid shell escape redirection problems
561 *) Initial support for certificate extension requests, these are included
562 in things like Xenroll certificate requests. Included functions to allow
563 extensions to be obtained and added.
566 *) -crlf option to s_client and s_server for sending newlines as
567 CRLF (as required by many protocols).
570 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
572 *) Install libRSAglue.a when OpenSSL is built with RSAref.
573 [Ralf S. Engelschall]
575 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
576 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
578 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
582 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
583 DH parameters/keys (q is lost during that conversion, but the resulting
584 DH parameters contain its length).
586 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
587 much faster than DH_generate_parameters (which creates parameters
588 where p = 2*q + 1), and also the smaller q makes DH computations
589 much more efficient (160-bit exponentiation instead of 1024-bit
590 exponentiation); so this provides a convenient way to support DHE
591 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
592 utter importance to use
593 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
595 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
596 when such DH parameters are used, because otherwise small subgroup
597 attacks may become possible!
600 *) Avoid memory leak in i2d_DHparams.
603 *) Allow the -k option to be used more than once in the enc program:
604 this allows the same encrypted message to be read by multiple recipients.
607 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
608 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
609 it will always use the numerical form of the OID, even if it has a short
613 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
614 method only got called if p,q,dmp1,dmq1,iqmp components were present,
615 otherwise bn_mod_exp was called. In the case of hardware keys for example
616 no private key components need be present and it might store extra data
617 in the RSA structure, which cannot be accessed from bn_mod_exp.
618 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
619 private key operations.
622 *) Added support for SPARC Linux.
625 *) pem_password_cb function type incompatibly changed from
626 typedef int pem_password_cb(char *buf, int size, int rwflag);
628 ....(char *buf, int size, int rwflag, void *userdata);
629 so that applications can pass data to their callbacks:
630 The PEM[_ASN1]_{read,write}... functions and macros now take an
631 additional void * argument, which is just handed through whenever
632 the password callback is called.
633 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
635 New function SSL_CTX_set_default_passwd_cb_userdata.
637 Compatibility note: As many C implementations push function arguments
638 onto the stack in reverse order, the new library version is likely to
639 interoperate with programs that have been compiled with the old
640 pem_password_cb definition (PEM_whatever takes some data that
641 happens to be on the stack as its last argument, and the callback
642 just ignores this garbage); but there is no guarantee whatsoever that
645 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
646 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
647 problems not only on Windows, but also on some Unix platforms.
648 To avoid problematic command lines, these definitions are now in an
649 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
650 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
653 *) MIPS III/IV assembler module is reimplemented.
656 *) More DES library cleanups: remove references to srand/rand and
657 delete an unused file.
660 *) Add support for the the free Netwide assembler (NASM) under Win32,
661 since not many people have MASM (ml) and it can be hard to obtain.
662 This is currently experimental but it seems to work OK and pass all
663 the tests. Check out INSTALL.W32 for info.
666 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
667 without temporary keys kept an extra copy of the server key,
668 and connections with temporary keys did not free everything in case
672 *) New function RSA_check_key and new openssl rsa option -check
673 for verifying the consistency of RSA keys.
674 [Ulf Moeller, Bodo Moeller]
676 *) Various changes to make Win32 compile work:
677 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
678 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
679 comparison" warnings.
680 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
683 *) Add a debugging option to PKCS#5 v2 key generation function: when
684 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
685 derived keys are printed to stderr.
688 *) Copy the flags in ASN1_STRING_dup().
689 [Roman E. Pavlov <pre@mo.msk.ru>]
691 *) The x509 application mishandled signing requests containing DSA
692 keys when the signing key was also DSA and the parameters didn't match.
694 It was supposed to omit the parameters when they matched the signing key:
695 the verifying software was then supposed to automatically use the CA's
696 parameters if they were absent from the end user certificate.
698 Omitting parameters is no longer recommended. The test was also
699 the wrong way round! This was probably due to unusual behaviour in
700 EVP_cmp_parameters() which returns 1 if the parameters match.
701 This meant that parameters were omitted when they *didn't* match and
702 the certificate was useless. Certificates signed with 'ca' didn't have
704 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
706 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
707 The interface is as follows:
709 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
710 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
711 "off" is now the default.
712 The library internally uses
713 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
714 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
715 to disable memory-checking temporarily.
717 Some inconsistent states that previously were possible (and were
718 even the default) are now avoided.
720 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
721 with each memory chunk allocated; this is occasionally more helpful
722 than just having a counter.
724 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
726 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
730 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
731 which largely parallels "options", but is for changing API behaviour,
732 whereas "options" are about protocol behaviour.
733 Initial "mode" flags are:
735 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
736 a single record has been written.
737 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
738 retries use the same buffer location.
739 (But all of the contents must be
743 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
746 *) Fix problems with no-hmac etc.
747 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
749 *) New functions RSA_get_default_method(), RSA_set_method() and
750 RSA_get_method(). These allows replacement of RSA_METHODs without having
751 to mess around with the internals of an RSA structure.
754 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
755 Also really enable memory leak checks in openssl.c and in some
757 [Chad C. Mulligan, Bodo Moeller]
759 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
760 up the length of negative integers. This has now been simplified to just
761 store the length when it is first determined and use it later, rather
762 than trying to keep track of where data is copied and updating it to
764 [Steve Henson, reported by Brien Wheeler
765 <bwheeler@authentica-security.com>]
767 *) Add a new function PKCS7_signatureVerify. This allows the verification
768 of a PKCS#7 signature but with the signing certificate passed to the
769 function itself. This contrasts with PKCS7_dataVerify which assumes the
770 certificate is present in the PKCS#7 structure. This isn't always the
771 case: certificates can be omitted from a PKCS#7 structure and be
772 distributed by "out of band" means (such as a certificate database).
775 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
776 function prototypes in pem.h, also change util/mkdef.pl to add the
777 necessary function names.
780 *) mk1mf.pl (used by Windows builds) did not properly read the
781 options set by Configure in the top level Makefile, and Configure
782 was not even able to write more than one option correctly.
783 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
786 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
787 file to be loaded from a BIO or FILE pointer. The BIO version will
788 for example allow memory BIOs to contain config info.
791 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
792 Whoever hopes to achieve shared-library compatibility across versions
793 must use this, not the compile-time macro.
794 (Exercise 0.9.4: Which is the minimum library version required by
796 Note: All this applies only to multi-threaded programs, others don't
800 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
801 through a BIO pair triggered the default case, i.e.
802 SSLerr(...,SSL_R_UNKNOWN_STATE).
805 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
806 can use the SSL library even if none of the specific BIOs is
810 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
811 for the encoded length.
812 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
814 *) Add initial documentation of the X509V3 functions.
817 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
818 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
819 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
820 secure PKCS#8 private key format with a high iteration count.
823 *) Fix determination of Perl interpreter: A perl or perl5
824 _directory_ in $PATH was also accepted as the interpreter.
825 [Ralf S. Engelschall]
827 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
828 wrong with it but it was very old and did things like calling
829 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
833 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
834 to use the new extension code.
837 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
838 with macros. This should make it easier to change their form, add extra
839 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
843 *) Add to configuration table a new entry that can specify an alternative
844 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
845 according to Mark Crispin <MRC@Panda.COM>.
849 *) DES CBC did not update the IV. Weird.
852 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
853 Changing the behaviour of the former might break existing programs --
854 where IV updating is needed, des_ncbc_encrypt can be used.
857 *) When bntest is run from "make test" it drives bc to check its
858 calculations, as well as internally checking them. If an internal check
859 fails, it needs to cause bc to give a non-zero result or make test carries
860 on without noticing the failure. Fixed.
863 *) DES library cleanups.
866 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
867 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
868 ciphers. NOTE: although the key derivation function has been verified
869 against some published test vectors it has not been extensively tested
870 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
874 *) Instead of "mkdir -p", which is not fully portable, use new
875 Perl script "util/mkdir-p.pl".
878 *) Rewrite the way password based encryption (PBE) is handled. It used to
879 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
880 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
881 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
882 the 'parameter' field of the AlgorithmIdentifier is passed to the
883 underlying key generation function so it must do its own ASN1 parsing.
884 This has also changed the EVP_PBE_CipherInit() function which now has a
885 'parameter' argument instead of literal salt and iteration count values
886 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
889 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
890 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
891 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
892 KEY" because this clashed with PKCS#8 unencrypted string. Since this
893 value was just used as a "magic string" and not used directly its
894 value doesn't matter.
897 *) Introduce some semblance of const correctness to BN. Shame C doesn't
901 *) "linux-sparc64" configuration (ultrapenguin).
902 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
903 "linux-sparc" configuration.
904 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
906 *) config now generates no-xxx options for missing ciphers.
909 *) Support the EBCDIC character set (work in progress).
910 File ebcdic.c not yet included because it has a different license.
911 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
913 *) Support BS2000/OSD-POSIX.
914 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
916 *) Make callbacks for key generation use void * instead of char *.
919 *) Make S/MIME samples compile (not yet tested).
922 *) Additional typesafe stacks.
925 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
929 Changes between 0.9.3 and 0.9.3a [29 May 1999]
931 *) New configuration variant "sco5-gcc".
933 *) Updated some demos.
934 [Sean O Riordain, Wade Scholine]
936 *) Add missing BIO_free at exit of pkcs12 application.
939 *) Fix memory leak in conf.c.
942 *) Updates for Win32 to assembler version of MD5.
945 *) Set #! path to perl in apps/der_chop to where we found it
946 instead of using a fixed path.
949 *) SHA library changes for irix64-mips4-cc.
952 *) Improvements for VMS support.
956 Changes between 0.9.2b and 0.9.3 [24 May 1999]
958 *) Bignum library bug fix. IRIX 6 passes "make test" now!
959 This also avoids the problems with SC4.2 and unpatched SC5.
960 [Andy Polyakov <appro@fy.chalmers.se>]
962 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
963 These are required because of the typesafe stack would otherwise break
964 existing code. If old code used a structure member which used to be STACK
965 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
966 sk_num or sk_value it would produce an error because the num, data members
967 are not present in STACK_OF. Now it just produces a warning. sk_set
968 replaces the old method of assigning a value to sk_value
969 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
970 that does this will no longer work (and should use sk_set instead) but
971 this could be regarded as a "questionable" behaviour anyway.
974 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
975 correctly handle encrypted S/MIME data.
978 *) Change type of various DES function arguments from des_cblock
979 (which means, in function argument declarations, pointer to char)
980 to des_cblock * (meaning pointer to array with 8 char elements),
981 which allows the compiler to do more typechecking; it was like
982 that back in SSLeay, but with lots of ugly casts.
984 Introduce new type const_des_cblock.
987 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
988 problems: find RecipientInfo structure that matches recipient certificate
989 and initialise the ASN1 structures properly based on passed cipher.
992 *) Belatedly make the BN tests actually check the results.
995 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
996 to and from BNs: it was completely broken. New compilation option
997 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
998 key elements as negative integers.
1001 *) Reorganize and speed up MD5.
1002 [Andy Polyakov <appro@fy.chalmers.se>]
1005 [Richard Levitte <richard@levitte.org>]
1007 *) New option -out to asn1parse to allow the parsed structure to be
1008 output to a file. This is most useful when combined with the -strparse
1009 option to examine the output of things like OCTET STRINGS.
1012 *) Make SSL library a little more fool-proof by not requiring any longer
1013 that SSL_set_{accept,connect}_state be called before
1014 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
1015 in many applications because usually everything *appeared* to work as
1016 intended anyway -- now it really works as intended).
1019 *) Move openssl.cnf out of lib/.
1022 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
1023 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
1024 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
1025 [Ralf S. Engelschall]
1027 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
1028 handle PKCS#7 enveloped data properly.
1029 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
1031 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
1032 copying pointers. The cert_st handling is changed by this in
1033 various ways (and thus what used to be known as ctx->default_cert
1034 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
1035 any longer when s->cert does not give us what we need).
1036 ssl_cert_instantiate becomes obsolete by this change.
1037 As soon as we've got the new code right (possibly it already is?),
1038 we have solved a couple of bugs of the earlier code where s->cert
1039 was used as if it could not have been shared with other SSL structures.
1041 Note that using the SSL API in certain dirty ways now will result
1042 in different behaviour than observed with earlier library versions:
1043 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
1044 does not influence s as it used to.
1046 In order to clean up things more thoroughly, inside SSL_SESSION
1047 we don't use CERT any longer, but a new structure SESS_CERT
1048 that holds per-session data (if available); currently, this is
1049 the peer's certificate chain and, for clients, the server's certificate
1050 and temporary key. CERT holds only those values that can have
1051 meaningful defaults in an SSL_CTX.
1054 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
1055 from the internal representation. Various PKCS#7 fixes: remove some
1056 evil casts and set the enc_dig_alg field properly based on the signing
1060 *) Allow PKCS#12 password to be set from the command line or the
1061 environment. Let 'ca' get its config file name from the environment
1062 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
1066 *) Allow certificate policies extension to use an IA5STRING for the
1067 organization field. This is contrary to the PKIX definition but
1068 VeriSign uses it and IE5 only recognises this form. Document 'x509'
1072 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
1073 without disallowing inline assembler and the like for non-pedantic builds.
1076 *) Support Borland C++ builder.
1077 [Janez Jere <jj@void.si>, modified by Ulf Möller]
1082 *) SHA-1 cleanups and performance enhancements.
1083 [Andy Polyakov <appro@fy.chalmers.se>]
1085 *) Sparc v8plus assembler for the bignum library.
1086 [Andy Polyakov <appro@fy.chalmers.se>]
1088 *) Accept any -xxx and +xxx compiler options in Configure.
1091 *) Update HPUX configuration.
1094 *) Add missing sk_<type>_unshift() function to safestack.h
1095 [Ralf S. Engelschall]
1097 *) New function SSL_CTX_use_certificate_chain_file that sets the
1098 "extra_cert"s in addition to the certificate. (This makes sense
1099 only for "PEM" format files, as chains as a whole are not
1103 *) Support verify_depth from the SSL API.
1104 x509_vfy.c had what can be considered an off-by-one-error:
1105 Its depth (which was not part of the external interface)
1106 was actually counting the number of certificates in a chain;
1107 now it really counts the depth.
1110 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
1111 instead of X509err, which often resulted in confusing error
1112 messages since the error codes are not globally unique
1113 (e.g. an alleged error in ssl3_accept when a certificate
1114 didn't match the private key).
1116 *) New function SSL_CTX_set_session_id_context that allows to set a default
1117 value (so that you don't need SSL_set_session_id_context for each
1118 connection using the SSL_CTX).
1121 *) OAEP decoding bug fix.
1124 *) Support INSTALL_PREFIX for package builders, as proposed by
1128 *) New Configure options "threads" and "no-threads". For systems
1129 where the proper compiler options are known (currently Solaris
1130 and Linux), "threads" is the default.
1133 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
1136 *) Install various scripts to $(OPENSSLDIR)/misc, not to
1137 $(INSTALLTOP)/bin -- they shouldn't clutter directories
1138 such as /usr/local/bin.
1141 *) "make linux-shared" to build shared libraries.
1142 [Niels Poppe <niels@netbox.org>]
1144 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
1147 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
1148 extension adding in x509 utility.
1151 *) Remove NOPROTO sections and error code comments.
1154 *) Partial rewrite of the DEF file generator to now parse the ANSI
1158 *) New Configure options --prefix=DIR and --openssldir=DIR.
1161 *) Complete rewrite of the error code script(s). It is all now handled
1162 by one script at the top level which handles error code gathering,
1163 header rewriting and C source file generation. It should be much better
1164 than the old method: it now uses a modified version of Ulf's parser to
1165 read the ANSI prototypes in all header files (thus the old K&R definitions
1166 aren't needed for error creation any more) and do a better job of
1167 translating function codes into names. The old 'ASN1 error code imbedded
1168 in a comment' is no longer necessary and it doesn't use .err files which
1169 have now been deleted. Also the error code call doesn't have to appear all
1170 on one line (which resulted in some large lines...).
1173 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
1176 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
1177 0 (which usually indicates a closed connection), but continue reading.
1180 *) Fix some race conditions.
1183 *) Add support for CRL distribution points extension. Add Certificate
1184 Policies and CRL distribution points documentation.
1187 *) Move the autogenerated header file parts to crypto/opensslconf.h.
1190 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
1191 8 of keying material. Merlin has also confirmed interop with this fix
1192 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
1193 [Merlin Hughes <merlin@baltimore.ie>]
1195 *) Fix lots of warnings.
1196 [Richard Levitte <levitte@stacken.kth.se>]
1198 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
1199 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
1200 [Richard Levitte <levitte@stacken.kth.se>]
1202 *) Fix problems with sizeof(long) == 8.
1203 [Andy Polyakov <appro@fy.chalmers.se>]
1205 *) Change functions to ANSI C.
1208 *) Fix typos in error codes.
1209 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
1211 *) Remove defunct assembler files from Configure.
1214 *) SPARC v8 assembler BIGNUM implementation.
1215 [Andy Polyakov <appro@fy.chalmers.se>]
1217 *) Support for Certificate Policies extension: both print and set.
1218 Various additions to support the r2i method this uses.
1221 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
1222 return a const string when you are expecting an allocated buffer.
1225 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
1226 types DirectoryString and DisplayText.
1229 *) Add code to allow r2i extensions to access the configuration database,
1230 add an LHASH database driver and add several ctx helper functions.
1233 *) Fix an evil bug in bn_expand2() which caused various BN functions to
1234 fail when they extended the size of a BIGNUM.
1237 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
1238 support typesafe stack.
1241 *) Fix typo in SSL_[gs]et_options().
1242 [Nils Frostberg <nils@medcom.se>]
1244 *) Delete various functions and files that belonged to the (now obsolete)
1245 old X509V3 handling code.
1248 *) New Configure option "rsaref".
1251 *) Don't auto-generate pem.h.
1254 *) Introduce type-safe ASN.1 SETs.
1257 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
1258 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
1260 *) Introduce type-safe STACKs. This will almost certainly break lots of code
1261 that links with OpenSSL (well at least cause lots of warnings), but fear
1262 not: the conversion is trivial, and it eliminates loads of evil casts. A
1263 few STACKed things have been converted already. Feel free to convert more.
1264 In the fullness of time, I'll do away with the STACK type altogether.
1267 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
1268 specified in <certfile> by updating the entry in the index.txt file.
1269 This way one no longer has to edit the index.txt file manually for
1270 revoking a certificate. The -revoke option does the gory details now.
1271 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
1273 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
1274 `-text' option at all and this way the `-noout -text' combination was
1275 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
1276 [Ralf S. Engelschall]
1278 *) Make sure a corresponding plain text error message exists for the
1279 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
1280 verify callback function determined that a certificate was revoked.
1281 [Ralf S. Engelschall]
1283 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
1284 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
1285 all available cipers including rc5, which was forgotten until now.
1286 In order to let the testing shell script know which algorithms
1287 are available, a new (up to now undocumented) command
1288 "openssl list-cipher-commands" is used.
1291 *) Bugfix: s_client occasionally would sleep in select() when
1292 it should have checked SSL_pending() first.
1295 *) New functions DSA_do_sign and DSA_do_verify to provide access to
1296 the raw DSA values prior to ASN.1 encoding.
1299 *) Tweaks to Configure
1300 [Niels Poppe <niels@netbox.org>]
1302 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
1306 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
1309 *) New config option to avoid instructions that are illegal on the 80386.
1310 The default code is faster, but requires at least a 486.
1313 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
1314 SSL2_SERVER_VERSION (not used at all) macros, which are now the
1315 same as SSL2_VERSION anyway.
1318 *) New "-showcerts" option for s_client.
1321 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
1322 application. Various cleanups and fixes.
1325 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
1326 modify error routines to work internally. Add error codes and PBE init
1327 to library startup routines.
1330 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
1331 packing functions to asn1 and evp. Changed function names and error
1332 codes along the way.
1335 *) PKCS12 integration: and so it begins... First of several patches to
1336 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
1337 objects to objects.h
1340 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
1341 and display support for Thawte strong extranet extension.
1344 *) Add LinuxPPC support.
1345 [Jeff Dubrule <igor@pobox.org>]
1347 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
1348 bn_div_words in alpha.s.
1349 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
1351 *) Make sure the RSA OAEP test is skipped under -DRSAref because
1352 OAEP isn't supported when OpenSSL is built with RSAref.
1353 [Ulf Moeller <ulf@fitug.de>]
1355 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
1356 so they no longer are missing under -DNOPROTO.
1357 [Soren S. Jorvang <soren@t.dk>]
1360 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
1362 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
1363 doesn't work when the session is reused. Coming soon!
1366 *) Fix a security hole, that allows sessions to be reused in the wrong
1367 context thus bypassing client cert protection! All software that uses
1368 client certs and session caches in multiple contexts NEEDS PATCHING to
1369 allow session reuse! A fuller solution is in the works.
1370 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
1372 *) Some more source tree cleanups (removed obsolete files
1373 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
1374 permission on "config" script to be executable) and a fix for the INSTALL
1376 [Ulf Moeller <ulf@fitug.de>]
1378 *) Remove some legacy and erroneous uses of malloc, free instead of
1380 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
1382 *) Make rsa_oaep_test return non-zero on error.
1383 [Ulf Moeller <ulf@fitug.de>]
1385 *) Add support for native Solaris shared libraries. Configure
1386 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
1387 if someone would make that last step automatic.
1388 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
1390 *) ctx_size was not built with the right compiler during "make links". Fixed.
1393 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
1394 except NULL ciphers". This means the default cipher list will no longer
1395 enable NULL ciphers. They need to be specifically enabled e.g. with
1396 the string "DEFAULT:eNULL".
1399 *) Fix to RSA private encryption routines: if p < q then it would
1400 occasionally produce an invalid result. This will only happen with
1401 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
1404 *) Be less restrictive and allow also `perl util/perlpath.pl
1405 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
1406 because this way one can also use an interpreter named `perl5' (which is
1407 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
1408 installed as `perl').
1409 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
1411 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
1412 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
1414 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
1415 advapi32.lib to Win32 build and change the pem test comparision
1416 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
1417 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
1418 and crypto/des/ede_cbcm_enc.c.
1421 *) DES quad checksum was broken on big-endian architectures. Fixed.
1424 *) Comment out two functions in bio.h that aren't implemented. Fix up the
1425 Win32 test batch file so it (might) work again. The Win32 test batch file
1426 is horrible: I feel ill....
1429 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
1430 in e_os.h. Audit of header files to check ANSI and non ANSI
1431 sections: 10 functions were absent from non ANSI section and not exported
1432 from Windows DLLs. Fixed up libeay.num for new functions.
1435 *) Make `openssl version' output lines consistent.
1436 [Ralf S. Engelschall]
1438 *) Fix Win32 symbol export lists for BIO functions: Added
1439 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
1440 to ms/libeay{16,32}.def.
1441 [Ralf S. Engelschall]
1443 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
1444 fine under Unix and passes some trivial tests I've now added. But the
1445 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
1446 added to make sure no one expects that this stuff really works in the
1447 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
1448 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
1450 [Ralf S. Engelschall]
1452 *) Fix the generation of two part addresses in perl.
1453 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
1455 *) Add config entry for Linux on MIPS.
1456 [John Tobey <jtobey@channel1.com>]
1458 *) Make links whenever Configure is run, unless we are on Windoze.
1461 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
1462 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
1466 *) Add a useful kludge to allow package maintainers to specify compiler and
1467 other platforms details on the command line without having to patch the
1468 Configure script everytime: One now can use ``perl Configure
1469 <id>:<details>'', i.e. platform ids are allowed to have details appended
1470 to them (seperated by colons). This is treated as there would be a static
1471 pre-configured entry in Configure's %table under key <id> with value
1472 <details> and ``perl Configure <id>'' is called. So, when you want to
1473 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
1474 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
1475 now, which overrides the FreeBSD-elf entry on-the-fly.
1476 [Ralf S. Engelschall]
1478 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
1481 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
1482 on the `perl Configure ...' command line. This way one can compile
1483 OpenSSL libraries with Position Independent Code (PIC) which is needed
1484 for linking it into DSOs.
1485 [Ralf S. Engelschall]
1487 *) Remarkably, export ciphers were totally broken and no-one had noticed!
1491 *) Cleaned up the LICENSE document: The official contact for any license
1492 questions now is the OpenSSL core team under openssl-core@openssl.org.
1493 And add a paragraph about the dual-license situation to make sure people
1494 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
1495 to the OpenSSL toolkit.
1496 [Ralf S. Engelschall]
1498 *) General source tree makefile cleanups: Made `making xxx in yyy...'
1499 display consistent in the source tree and replaced `/bin/rm' by `rm'.
1500 Additonally cleaned up the `make links' target: Remove unnecessary
1501 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
1502 to speed processing and no longer clutter the display with confusing
1503 stuff. Instead only the actually done links are displayed.
1504 [Ralf S. Engelschall]
1506 *) Permit null encryption ciphersuites, used for authentication only. It used
1507 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
1508 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
1512 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
1513 signed attributes when verifying signatures (this would break them),
1514 the detached data encoding was wrong and public keys obtained using
1515 X509_get_pubkey() weren't freed.
1518 *) Add text documentation for the BUFFER functions. Also added a work around
1519 to a Win95 console bug. This was triggered by the password read stuff: the
1520 last character typed gets carried over to the next fread(). If you were
1521 generating a new cert request using 'req' for example then the last
1522 character of the passphrase would be CR which would then enter the first
1526 *) Added the new `Includes OpenSSL Cryptography Software' button as
1527 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
1528 button and can be used by applications based on OpenSSL to show the
1529 relationship to the OpenSSL project.
1530 [Ralf S. Engelschall]
1532 *) Remove confusing variables in function signatures in files
1533 ssl/ssl_lib.c and ssl/ssl.h.
1534 [Lennart Bong <lob@kulthea.stacken.kth.se>]
1536 *) Don't install bss_file.c under PREFIX/include/
1537 [Lennart Bong <lob@kulthea.stacken.kth.se>]
1539 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
1540 functions that return function pointers and has support for NT specific
1541 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
1542 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
1543 unsigned to signed types: this was killing the Win32 compile.
1546 *) Add new certificate file to stack functions,
1547 SSL_add_dir_cert_subjects_to_stack() and
1548 SSL_add_file_cert_subjects_to_stack(). These largely supplant
1549 SSL_load_client_CA_file(), and can be used to add multiple certs easily
1550 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
1551 This means that Apache-SSL and similar packages don't have to mess around
1552 to add as many CAs as they want to the preferred list.
1555 *) Experiment with doxygen documentation. Currently only partially applied to
1557 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
1558 openssl.doxy as the configuration file.
1561 *) Get rid of remaining C++-style comments which strict C compilers hate.
1562 [Ralf S. Engelschall, pointed out by Carlos Amengual]
1564 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
1565 compiled in by default: it has problems with large keys.
1568 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
1569 DH private keys and/or callback functions which directly correspond to
1570 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
1571 is needed for applications which have to configure certificates on a
1572 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
1574 For the RSA certificate situation is makes no difference, but
1575 for the DSA certificate situation this fixes the "no shared cipher"
1576 problem where the OpenSSL cipher selection procedure failed because the
1577 temporary keys were not overtaken from the context and the API provided
1578 no way to reconfigure them.
1579 The new functions now let applications reconfigure the stuff and they
1580 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
1581 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
1582 non-public-API function ssl_cert_instantiate() is used as a helper
1583 function and also to reduce code redundancy inside ssl_rsa.c.
1584 [Ralf S. Engelschall]
1586 *) Move s_server -dcert and -dkey options out of the undocumented feature
1587 area because they are useful for the DSA situation and should be
1588 recognized by the users.
1589 [Ralf S. Engelschall]
1591 *) Fix the cipher decision scheme for export ciphers: the export bits are
1592 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
1593 SSL_EXP_MASK. So, the original variable has to be used instead of the
1594 already masked variable.
1595 [Richard Levitte <levitte@stacken.kth.se>]
1597 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
1598 [Richard Levitte <levitte@stacken.kth.se>]
1600 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
1601 from `int' to `unsigned int' because it's a length and initialized by
1602 EVP_DigestFinal() which expects an `unsigned int *'.
1603 [Richard Levitte <levitte@stacken.kth.se>]
1605 *) Don't hard-code path to Perl interpreter on shebang line of Configure
1606 script. Instead use the usual Shell->Perl transition trick.
1607 [Ralf S. Engelschall]
1609 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
1610 (in addition to RSA certificates) to match the behaviour of `openssl dsa
1611 -noout -modulus' as it's already the case for `openssl rsa -noout
1612 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
1613 currently the public key is printed (a decision which was already done by
1614 `openssl dsa -modulus' in the past) which serves a similar purpose.
1615 Additionally the NO_RSA no longer completely removes the whole -modulus
1616 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
1618 [Ralf S. Engelschall]
1620 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
1621 BIO. See the source (crypto/evp/bio_ok.c) for more info.
1622 [Arne Ansper <arne@ats.cyber.ee>]
1624 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
1625 to be added. Now both 'req' and 'ca' can use new objects defined in the
1629 *) Add cool BIO that does syslog (or event log on NT).
1630 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
1632 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
1633 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
1634 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
1635 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
1638 *) Add preliminary config info for new extension code.
1641 *) Make RSA_NO_PADDING really use no padding.
1642 [Ulf Moeller <ulf@fitug.de>]
1644 *) Generate errors when private/public key check is done.
1647 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
1648 for some CRL extensions and new objects added.
1651 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
1652 key usage extension and fuller support for authority key id.
1655 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
1656 padding method for RSA, which is recommended for new applications in PKCS
1657 #1 v2.0 (RFC 2437, October 1998).
1658 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
1659 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
1660 against Bleichbacher's attack on RSA.
1661 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
1664 *) Updates to the new SSL compression code
1665 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
1667 *) Fix so that the version number in the master secret, when passed
1668 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
1669 (because the server will not accept higher), that the version number
1670 is 0x03,0x01, not 0x03,0x00
1671 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
1673 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
1674 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
1675 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
1678 *) Support for RAW extensions where an arbitrary extension can be
1679 created by including its DER encoding. See apps/openssl.cnf for
1683 *) Make sure latest Perl versions don't interpret some generated C array
1684 code as Perl array code in the crypto/err/err_genc.pl script.
1685 [Lars Weber <3weber@informatik.uni-hamburg.de>]
1687 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
1688 not many people have the assembler. Various Win32 compilation fixes and
1689 update to the INSTALL.W32 file with (hopefully) more accurate Win32
1693 *) Modify configure script 'Configure' to automatically create crypto/date.h
1694 file under Win32 and also build pem.h from pem.org. New script
1695 util/mkfiles.pl to create the MINFO file on environments that can't do a
1696 'make files': perl util/mkfiles.pl >MINFO should work.
1699 *) Major rework of DES function declarations, in the pursuit of correctness
1700 and purity. As a result, many evil casts evaporated, and some weirdness,
1701 too. You may find this causes warnings in your code. Zapping your evil
1702 casts will probably fix them. Mostly.
1705 *) Fix for a typo in asn1.h. Bug fix to object creation script
1706 obj_dat.pl. It considered a zero in an object definition to mean
1707 "end of object": none of the objects in objects.h have any zeros
1708 so it wasn't spotted.
1709 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
1711 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
1712 Masking (CBCM). In the absence of test vectors, the best I have been able
1713 to do is check that the decrypt undoes the encrypt, so far. Send me test
1714 vectors if you have them.
1717 *) Correct calculation of key length for export ciphers (too much space was
1718 allocated for null ciphers). This has not been tested!
1721 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
1722 message is now correct (it understands "crypto" and "ssl" on its
1723 command line). There is also now an "update" option. This will update
1724 the util/ssleay.num and util/libeay.num files with any new functions.
1726 perl util/mkdef.pl crypto ssl update
1727 it will update them.
1730 *) Overhauled the Perl interface (perl/*):
1731 - ported BN stuff to OpenSSL's different BN library
1732 - made the perl/ source tree CVS-aware
1733 - renamed the package from SSLeay to OpenSSL (the files still contain
1734 their history because I've copied them in the repository)
1735 - removed obsolete files (the test scripts will be replaced
1736 by better Test::Harness variants in the future)
1737 [Ralf S. Engelschall]
1739 *) First cut for a very conservative source tree cleanup:
1740 1. merge various obsolete readme texts into doc/ssleay.txt
1741 where we collect the old documents and readme texts.
1742 2. remove the first part of files where I'm already sure that we no
1743 longer need them because of three reasons: either they are just temporary
1744 files which were left by Eric or they are preserved original files where
1745 I've verified that the diff is also available in the CVS via "cvs diff
1746 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
1747 the crypto/md/ stuff).
1748 [Ralf S. Engelschall]
1750 *) More extension code. Incomplete support for subject and issuer alt
1751 name, issuer and authority key id. Change the i2v function parameters
1752 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
1753 what that's for :-) Fix to ASN1 macro which messed up
1754 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
1757 *) Preliminary support for ENUMERATED type. This is largely copied from the
1761 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
1762 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
1764 *) Make sure `make rehash' target really finds the `openssl' program.
1765 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
1767 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
1768 like to hear about it if this slows down other processors.
1771 *) Add CygWin32 platform information to Configure script.
1772 [Alan Batie <batie@aahz.jf.intel.com>]
1774 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
1775 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
1777 *) New program nseq to manipulate netscape certificate sequences
1780 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
1784 *) Fixes to BN code. Previously the default was to define BN_RECURSION
1785 but the BN code had some problems that would cause failures when
1786 doing certificate verification and some other functions.
1787 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
1789 *) Add ASN1 and PEM code to support netscape certificate sequences.
1792 *) Add ASN1 and PEM code to support netscape certificate sequences.
1795 *) Add several PKIX and private extended key usage OIDs.
1798 *) Modify the 'ca' program to handle the new extension code. Modify
1799 openssl.cnf for new extension format, add comments.
1802 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
1803 and add a sample to openssl.cnf so req -x509 now adds appropriate
1807 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
1808 error code, add initial support to X509_print() and x509 application.
1811 *) Takes a deep breath and start addding X509 V3 extension support code. Add
1812 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
1813 stuff is currently isolated and isn't even compiled yet.
1816 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
1817 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
1818 Removed the versions check from X509 routines when loading extensions:
1819 this allows certain broken certificates that don't set the version
1820 properly to be processed.
1823 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
1824 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
1825 can still be regenerated with "make depend".
1828 *) Spelling mistake in C version of CAST-128.
1829 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
1831 *) Changes to the error generation code. The perl script err-code.pl
1832 now reads in the old error codes and retains the old numbers, only
1833 adding new ones if necessary. It also only changes the .err files if new
1834 codes are added. The makefiles have been modified to only insert errors
1835 when needed (to avoid needlessly modifying header files). This is done
1836 by only inserting errors if the .err file is newer than the auto generated
1837 C file. To rebuild all the error codes from scratch (the old behaviour)
1838 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
1839 or delete all the .err files.
1842 *) CAST-128 was incorrectly implemented for short keys. The C version has
1843 been fixed, but is untested. The assembler versions are also fixed, but
1844 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
1845 to regenerate it if needed.
1846 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
1847 Hagino <itojun@kame.net>]
1849 *) File was opened incorrectly in randfile.c.
1850 [Ulf Möller <ulf@fitug.de>]
1852 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
1853 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
1854 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
1855 al: it's just almost always a UTCTime. Note this patch adds new error
1856 codes so do a "make errors" if there are problems.
1859 *) Correct Linux 1 recognition in config.
1860 [Ulf Möller <ulf@fitug.de>]
1862 *) Remove pointless MD5 hash when using DSA keys in ca.
1863 [Anonymous <nobody@replay.com>]
1865 *) Generate an error if given an empty string as a cert directory. Also
1866 generate an error if handed NULL (previously returned 0 to indicate an
1867 error, but didn't set one).
1868 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
1870 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
1873 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
1874 parameters. This was causing a warning which killed off the Win32 compile.
1877 *) Remove C++ style comments from crypto/bn/bn_local.h.
1878 [Neil Costigan <neil.costigan@celocom.com>]
1880 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
1881 based on a text string, looking up short and long names and finally
1882 "dot" format. The "dot" format stuff didn't work. Added new function
1883 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
1884 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
1885 OID is not part of the table.
1888 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
1889 X509_LOOKUP_by_alias().
1892 *) Sort openssl functions by name.
1895 *) Get the gendsa program working (hopefully) and add it to app list. Remove
1896 encryption from sample DSA keys (in case anyone is interested the password
1900 *) Make _all_ *_free functions accept a NULL pointer.
1901 [Frans Heymans <fheymans@isaserver.be>]
1903 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
1905 [Anonymous <nobody@replay.com>]
1907 *) s_server should send the CAfile as acceptable CAs, not its own cert.
1908 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
1910 *) Don't blow it for numeric -newkey arguments to apps/req.
1911 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
1913 *) Temp key "for export" tests were wrong in s3_srvr.c.
1914 [Anonymous <nobody@replay.com>]
1916 *) Add prototype for temp key callback functions
1917 SSL_CTX_set_tmp_{rsa,dh}_callback().
1920 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
1921 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
1924 *) X509_name_add_entry() freed the wrong thing after an error.
1925 [Arne Ansper <arne@ats.cyber.ee>]
1927 *) rsa_eay.c would attempt to free a NULL context.
1928 [Arne Ansper <arne@ats.cyber.ee>]
1930 *) BIO_s_socket() had a broken should_retry() on Windoze.
1931 [Arne Ansper <arne@ats.cyber.ee>]
1933 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
1934 [Arne Ansper <arne@ats.cyber.ee>]
1936 *) Make sure the already existing X509_STORE->depth variable is initialized
1937 in X509_STORE_new(), but document the fact that this variable is still
1938 unused in the certificate verification process.
1939 [Ralf S. Engelschall]
1941 *) Fix the various library and apps files to free up pkeys obtained from
1942 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
1945 *) Fix reference counting in X509_PUBKEY_get(). This makes
1946 demos/maurice/example2.c work, amongst others, probably.
1947 [Steve Henson and Ben Laurie]
1949 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
1950 `openssl' and second, the shortcut symlinks for the `openssl <command>'
1951 are no longer created. This way we have a single and consistent command
1952 line interface `openssl <command>', similar to `cvs <command>'.
1953 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
1955 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
1956 BIT STRING wrapper always have zero unused bits.
1959 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
1962 *) Make the top-level INSTALL documentation easier to understand.
1965 *) Makefiles updated to exit if an error occurs in a sub-directory
1966 make (including if user presses ^C) [Paul Sutton]
1968 *) Make Montgomery context stuff explicit in RSA data structure.
1971 *) Fix build order of pem and err to allow for generated pem.h.
1974 *) Fix renumbering bug in X509_NAME_delete_entry().
1977 *) Enhanced the err-ins.pl script so it makes the error library number
1978 global and can add a library name. This is needed for external ASN1 and
1979 other error libraries.
1982 *) Fixed sk_insert which never worked properly.
1985 *) Fix ASN1 macros so they can handle indefinite length construted
1986 EXPLICIT tags. Some non standard certificates use these: they can now
1990 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
1991 into a single doc/ssleay.txt bundle. This way the information is still
1992 preserved but no longer messes up this directory. Now it's new room for
1993 the new set of documenation files.
1994 [Ralf S. Engelschall]
1996 *) SETs were incorrectly DER encoded. This was a major pain, because they
1997 shared code with SEQUENCEs, which aren't coded the same. This means that
1998 almost everything to do with SETs or SEQUENCEs has either changed name or
1999 number of arguments.
2000 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
2002 *) Fix test data to work with the above.
2005 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
2006 was already fixed by Eric for 0.9.1 it seems.
2007 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
2009 *) Autodetect FreeBSD3.
2012 *) Fix various bugs in Configure. This affects the following platforms:
2016 unixware-2.0-pentium
2020 *) Eliminate generated files from CVS. Reorder tests to regenerate files
2021 before they are needed.
2024 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
2028 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
2030 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
2031 changed SSLeay to OpenSSL in version strings.
2032 [Ralf S. Engelschall]
2034 *) Some fixups to the top-level documents.
2037 *) Fixed the nasty bug where rsaref.h was not found under compile-time
2038 because the symlink to include/ was missing.
2039 [Ralf S. Engelschall]
2041 *) Incorporated the popular no-RSA/DSA-only patches
2042 which allow to compile a RSA-free SSLeay.
2043 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
2045 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
2046 when "ssleay" is still not found.
2047 [Ralf S. Engelschall]
2049 *) Added more platforms to Configure: Cray T3E, HPUX 11,
2050 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
2052 *) Updated the README file.
2053 [Ralf S. Engelschall]
2055 *) Added various .cvsignore files in the CVS repository subdirs
2056 to make a "cvs update" really silent.
2057 [Ralf S. Engelschall]
2059 *) Recompiled the error-definition header files and added
2060 missing symbols to the Win32 linker tables.
2061 [Ralf S. Engelschall]
2063 *) Cleaned up the top-level documents;
2064 o new files: CHANGES and LICENSE
2065 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
2066 o merged COPYRIGHT into LICENSE
2067 o removed obsolete TODO file
2068 o renamed MICROSOFT to INSTALL.W32
2069 [Ralf S. Engelschall]
2071 *) Removed dummy files from the 0.9.1b source tree:
2072 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
2073 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
2074 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
2075 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
2076 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
2077 [Ralf S. Engelschall]
2079 *) Added various platform portability fixes.
2082 *) The Genesis of the OpenSSL rpject:
2083 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
2084 Young and Tim J. Hudson created while they were working for C2Net until
2086 [The OpenSSL Project]
2089 Changes between 0.9.0b and 0.9.1b [not released]
2091 *) Updated a few CA certificates under certs/
2094 *) Changed some BIGNUM api stuff.
2097 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
2098 DGUX x86, Linux Alpha, etc.
2101 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
2102 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
2106 *) Add -strparse option to asn1pars program which parses nested
2108 [Dr Stephen Henson <shenson@bigfoot.com>]
2110 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
2113 *) DSA fix for "ca" program.
2116 *) Added "-genkey" option to "dsaparam" program.
2119 *) Added RIPE MD160 (rmd160) message digest.
2122 *) Added -a (all) option to "ssleay version" command.
2125 *) Added PLATFORM define which is the id given to Configure.
2128 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
2131 *) Extended the ASN.1 parser routines.
2134 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
2137 *) Added a BN_CTX to the BN library.
2140 *) Fixed the weak key values in DES library
2143 *) Changed API in EVP library for cipher aliases.
2146 *) Added support for RC2/64bit cipher.
2149 *) Converted the lhash library to the crypto/mem.c functions.
2152 *) Added more recognized ASN.1 object ids.
2155 *) Added more RSA padding checks for SSL/TLS.
2158 *) Added BIO proxy/filter functionality.
2161 *) Added extra_certs to SSL_CTX which can be used
2162 send extra CA certificates to the client in the CA cert chain sending
2163 process. It can be configured with SSL_CTX_add_extra_chain_cert().
2166 *) Now Fortezza is denied in the authentication phase because
2167 this is key exchange mechanism is not supported by SSLeay at all.
2170 *) Additional PKCS1 checks.
2173 *) Support the string "TLSv1" for all TLS v1 ciphers.
2176 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
2177 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
2180 *) Fixed a few memory leaks.
2183 *) Fixed various code and comment typos.
2186 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
2187 bytes sent in the client random.
2188 [Edward Bishop <ebishop@spyglass.com>]