4 Changes between 0.9.6 and 0.9.7 [xx XXX 2000]
6 *) Fix bug in PKCS#12 key generation routines. This was triggered
7 if a 3DES key was generated with a 0 initial byte. Include
8 PKCS12_BROKEN_KEYGEN compilation option to retain the old
9 (but broken) behaviour.
12 *) Enhance bctest to search for a working bc along $PATH and print
14 [Tim Rice <tim@multitalents.net> via Richard Levitte]
16 *) Add a 'copy_extensions' option to the 'ca' utility. This copies
17 extensions from a certificate request to the certificate.
20 *) Allow multiple 'certopt' and 'nameopt' options to be separated
21 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
22 file: this allows the display of the certificate about to be
23 signed to be customised, to allow certain fields to be included
24 or excluded and extension details. The old system didn't display
25 multicharacter strings properly, omitted fields not in the policy
26 and couldn't display additional details such as extensions.
29 *) Fix memory leaks in err.c: free err_data string if necessary;
30 don't write to the wrong index in ERR_set_error_data.
33 *) Function EC_POINTs_mul for simultaneous scalar multiplication
34 of an arbitrary number of elliptic curve points, optionally
35 including the generator defined for the EC_GROUP.
36 EC_POINT_mul is a simple wrapper function for the typical case
37 that the point list has just one item (besides the optional
41 *) First EC_METHODs for curves over GF(p):
43 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
44 operations and provides various method functions that can also
45 operate with faster implementations of modular arithmetic.
47 EC_GFp_mont_method() reuses most functions that are part of
48 EC_GFp_simple_method, but uses Montgomery arithmetic.
50 [Bodo Moeller; point addition and point doubling
51 implementation directly derived from source code provided by
52 Lenka Fibikova <fibikova@exp-math.uni-essen.de>]
54 *) Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
57 Curves are EC_GROUP objects (with an optional group generator)
58 based on EC_METHODs that are built into the library.
60 Points are EC_POINT objects based on EC_GROUP objects.
62 Most of the framework would be able to handle curves over arbitrary
63 finite fields, but as there are no obvious types for fields other
64 than GF(p), some functions are limited to that for now.
67 *) Add the -HTTP option to s_server. It is similar to -WWW, but requires
68 that the file contains a complete HTTP response.
71 *) Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
72 change the def and num file printf format specifier from "%-40sXXX"
73 to "%-39s XXX". The latter will always guarantee a space after the
74 field while the former will cause them to run together if the field
75 is 40 of more characters long.
78 *) Constify the cipher and digest 'method' functions and structures
79 and modify related functions to take constant EVP_MD and EVP_CIPHER
83 *) Implement ssl23_peek (analogous to ssl23_read), which previously
87 *) Replace rdtsc with _emit statements for VC++ version 5.
88 [Jeremy Cooper <jeremy@baymoo.org>]
90 *) Hide BN_CTX structure details in bn_lcl.h instead of publishing them
91 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
94 *) Modify EVP_Digest*() routines so they now return values. Although the
95 internal software routines can never fail additional hardware versions
99 *) Clean up crypto/err/err.h and change some error codes to avoid conflicts:
101 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
102 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
105 ERR_R_NESTED_ASN1_ERROR
107 ERR_R_MISSING_ASN1_EOS
108 were 4 .. 9, conflicting with
109 ERR_LIB_RSA (= ERR_R_RSA_LIB)
111 ERR_LIB_PEM (= ERR_R_PEM_LIB).
112 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
114 Add new error code 'ERR_R_INTERNAL_ERROR'.
117 *) Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
121 *) New option '-subj arg' for 'openssl req' and 'openssl ca'. This
122 sets the subject name for a new request or supersedes the
123 subject name in a given request. Formats that can be parsed are
124 'CN=Some Name, OU=myOU, C=IT'
126 'CN=Some Name/OU=myOU/C=IT'.
128 Add options '-batch' and '-verbose' to 'openssl req'.
129 [Massimiliano Pala <madwolf@hackmasters.net>]
131 *) Make it possible to reuse SSLv2 sessions.
134 *) Introduce the possibility to access global variables through
135 functions on platform were that's the best way to handle exporting
136 global variables in shared libraries. To enable this functionality,
137 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
138 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
139 is normally done by Configure or something similar).
141 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
142 in the source file (foo.c) like this:
144 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
145 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
147 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
148 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
150 OPENSSL_DECLARE_GLOBAL(int,foo);
151 #define foo OPENSSL_GLOBAL_REF(foo)
152 OPENSSL_DECLARE_GLOBAL(double,bar);
153 #define bar OPENSSL_GLOBAL_REF(bar)
155 The #defines are very important, and therefore so is including the
156 header file everywere where the defined globals are used.
158 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
159 of ASN.1 items, but that structure is a bt different.
161 The largest change is in util/mkdef.pl which has been enhanced with
162 better and easier to understand logic to choose which symbols should
163 go into the Windows .def files as well as a number of fixes and code
164 cleanup (among others, algorithm keywords are now sorted
165 lexicographically to avoid constant rewrites).
168 *) In copy_email() check for >= 0 as a return value for
169 X509_NAME_get_index_by_NID() since 0 is a valid index.
170 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>]
172 *) In BN_div() keep a copy of the sign of 'num' before writing the
173 result to 'rm' because if rm==num the value will be overwritten
174 and produce the wrong result if 'num' is negative: this caused
175 problems with BN_mod() and BN_nnmod().
178 *) Function OCSP_request_verify(). This checks the signature on an
179 OCSP request and verifies the signer certificate. The signer
180 certificate is just checked for a generic purpose and OCSP request
184 *) Add OCSP_check_validity() function to check the validity of OCSP
185 responses. OCSP responses are prepared in real time and may only
186 be a few seconds old. Simply checking that the current time lies
187 between thisUpdate and nextUpdate max reject otherwise valid responses
188 caused by either OCSP responder or client clock innacuracy. Instead
189 we allow thisUpdate and nextUpdate to fall within a certain period of
190 the current time. The age of the response can also optionally be
191 checked. Two new options -validity_period and -status_age added to
195 *) If signature or public key algorithm is unrecognized print out its
196 OID rather that just UNKOWN.
199 *) Avoid coredump with unsupported or invalid public keys by checking if
200 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
201 PKCS7_verify() fails with non detached data.
204 *) Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
205 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
206 ID to be generated from the issuer certificate alone which can then be
207 passed to OCSP_id_issuer_cmp().
210 *) New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
211 ASN1 modules to export functions returning ASN1_ITEM pointers
212 instead of the ASN1_ITEM structures themselves. This adds several
213 new macros which allow the underlying ASN1 function/structure to
214 be accessed transparently. As a result code should not use ASN1_ITEM
215 references directly (such as &X509_it) but instead use the relevant
216 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
217 use of the new ASN1 code on platforms where exporting structures
218 is problematical (for example in shared libraries) but exporting
219 functions returning pointers to structures is not.
222 *) Add support for overriding the generation of SSL/TLS session IDs.
223 These callbacks can be registered either in an SSL_CTX or per SSL.
224 The purpose of this is to allow applications to control, if they wish,
225 the arbitrary values chosen for use as session IDs, particularly as it
226 can be useful for session caching in multiple-server environments. A
227 command-line switch for testing this (and any client code that wishes
228 to use such a feature) has been added to "s_server".
229 [Geoff Thorpe, Lutz Jaenicke]
231 *) Modify mkdef.pl to recognise and parse prprocessor conditionals
232 of the form '#if defined(...) || defined(...) || ...' and
233 '#if !defined(...) && !defined(...) && ...'. This also avoids
234 the growing number of special cases it was previously handling.
237 *) Don't use getenv in library functions when run as setuid/setgid.
238 New function OPENSSL_issetugid().
241 *) Make all configuration macros available for application by making
242 sure they are available in opensslconf.h, by giving them names starting
243 with "OPENSSL_" to avoid conflicts with other packages and by making
244 sure e_os2.h will cover all platform-specific cases together with
246 Additionally, it is now possible to define configuration/platform-
247 specific names (called "system identities"). In the C code, these
248 are prefixed with "OPENSSL_SYSNAME_". e_os2.h will create another
249 macro with the name beginning with "OPENSSL_SYS_", which is determined
250 from "OPENSSL_SYSNAME_*" or compiler-specific macros depending on
254 *) New option -set_serial to 'req' and 'x509' this allows the serial
255 number to use to be specified on the command line. Previously self
256 signed certificates were hard coded with serial number 0 and the
257 CA options of 'x509' had to use a serial number in a file which was
261 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
262 due to incorrect handling of multi-threading:
264 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
266 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
268 3. Count how many times MemCheck_off() has been called so that
269 nested use can be treated correctly. This also avoids
270 inband-signalling in the previous code (which relied on the
271 assumption that thread ID 0 is impossible).
274 *) New options to 'ca' utility to support V2 CRL entry extensions.
275 Currently CRL reason, invalidity date and hold instruction are
276 supported. Add new CRL extensions to V3 code and some new objects.
279 *) Add "-rand" option also to s_client and s_server.
282 *) New function EVP_CIPHER_CTX_set_padding() this is used to
283 disable standard block padding (aka PKCS#5 padding) in the EVP
284 API, which was previously mandatory. This means that the data is
285 not padded in any way and so the total length much be a multiple
286 of the block size, otherwise an error occurs.
289 *) Initial (incomplete) OCSP SSL support.
292 *) Fix CPU detection on Irix 6.x.
293 [Kurt Hockenbury <khockenb@stevens-tech.edu> and
294 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
296 *) New function OCSP_parse_url(). This splits up a URL into its host,
297 port and path components: primarily to parse OCSP URLs. New -url
298 option to ocsp utility.
301 *) New nonce behavior. The return value of OCSP_check_nonce() now
302 reflects the various checks performed. Applications can decide
303 whether to tolerate certain situations such as an absent nonce
304 in a response when one was present in a request: the ocsp application
305 just prints out a warning. New function OCSP_add1_basic_nonce()
306 this is to allow responders to include a nonce in a response even if
307 the request is nonce-less.
310 *) Use the cached encoding of an X509_NAME structure rather than
311 copying it. This is apparently the reason for the libsafe "errors"
312 but the code is actually correct.
315 *) Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
316 skipped when using openssl x509 multiple times on a single input file,
317 e.g. "(openssl x509 -out cert1; openssl x509 -out cert2) <certs".
320 *) Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
321 set string type: to handle setting ASN1_TIME structures. Fix ca
322 utility to correctly initialize revocation date of CRLs.
325 *) New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
326 the clients preferred ciphersuites and rather use its own preferences.
327 Should help to work around M$ SGC (Server Gated Cryptography) bug in
328 Internet Explorer by ensuring unchanged hash method during stepup.
331 *) Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
332 to aes and add a new 'exist' option to print out symbols that don't
336 *) Additional options to ocsp utility to allow flags to be set and
337 additional certificates supplied.
340 *) Add the option -VAfile to 'openssl ocsp', so the user can give the
341 OCSP client a number of certificate to only verify the response
345 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
346 Bleichenbacher's DSA attack.
347 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
348 to be set and top=0 forces the highest bit to be set; top=-1 is new
349 and leaves the highest bit random.
350 [Ulf Moeller, Bodo Moeller]
352 *) Update Rijndael code to version 3.0 and change EVP AES ciphers to
353 handle the new API. Currently only ECB, CBC modes supported. Add new
354 AES OIDs. Add TLS AES ciphersuites as described in the "AES Ciphersuites
355 for TLS" draft-ietf-tls-ciphersuite-03.txt.
356 [Ben Laurie, Steve Henson]
358 *) In the NCONF_...-based implementations for CONF_... queries
359 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
360 a temporary CONF structure with the data component set to NULL
361 (which gives segmentation faults in lh_retrieve).
362 Instead, use NULL for the CONF pointer in CONF_get_string and
363 CONF_get_number (which may use environment variables) and directly
364 return NULL from CONF_get_section.
367 *) Fix potential buffer overrun for EBCDIC.
370 *) New function OCSP_copy_nonce() to copy nonce value (if present) from
374 *) Functions for OCSP responders. OCSP_request_onereq_count(),
375 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
376 extract information from a certificate request. OCSP_response_create()
377 creates a response and optionally adds a basic response structure.
378 OCSP_basic_add1_status() adds a complete single response to a basic
379 reponse and returns the OCSP_SINGLERESP structure just added (to allow
380 extensions to be included for example). OCSP_basic_add1_cert() adds a
381 certificate to a basic response and OCSP_basic_sign() signs a basic
382 response with various flags. New helper functions ASN1_TIME_check()
383 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
384 (converts ASN1_TIME to GeneralizedTime).
387 *) Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
388 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
389 structure from a certificate. X509_pubkey_digest() digests tha public_key
390 contents: this is used in various key identifiers.
393 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign
394 keyUsage if basicConstraints absent for a CA.
397 *) Make SMIME_write_PKCS7() write mail header values with a format that
398 is more generally accepted (no spaces before the semicolon), since
399 some programs can't parse those values properly otherwise. Also make
400 sure BIO's that break lines after each write do not create invalid
404 *) Make sk_sort() tolerate a NULL argument.
405 [Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>]
407 *) New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
408 passed by the function are trusted implicitly. If any of them signed the
409 reponse then it is assumed to be valid and is not verified.
412 *) Zero the premaster secret after deriving the master secret in
416 *) In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
417 to data. This was previously part of the PKCS7 ASN1 code. This
418 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
419 [Steve Henson, reported by Kenneth R. Robinette
420 <support@securenetterm.com>]
422 *) Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
423 routines: without these tracing memory leaks is very painful.
424 Fix leaks in PKCS12 and PKCS7 routines.
427 *) Fix for Irix with NO_ASM.
428 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
430 *) Add some EVP_add_digest_alias registrations (as found in
431 OpenSSL_add_all_digests()), to SSL_library_init()
432 aka OpenSSL_add_ssl_algorithms(). This provides improved
433 compatibility with peers using X.509 certificates
434 with unconventional AlgorithmIdentifier OIDs.
437 *) ./config script fixes.
438 [Ulf Moeller, Richard Levitte]
440 *) Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
441 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
442 effectively meant GeneralizedTime would never be used. Now it
443 is initialised to -1 but X509_time_adj() now has to check the value
444 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
445 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
446 [Steve Henson, reported by Kenneth R. Robinette
447 <support@securenetterm.com>]
449 *) Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
450 result in a zero length in the ASN1_INTEGER structure which was
451 not consistent with the structure when d2i_ASN1_INTEGER() was used
452 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
453 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
454 where it did not print out a minus for negative ASN1_INTEGER.
457 *) Fix 'openssl passwd -1'.
460 *) Add summary printout to ocsp utility. The various functions which
461 convert status values to strings have been renamed to:
462 OCSP_response_status_str(), OCSP_cert_status_str() and
463 OCSP_crl_reason_str() and are no longer static. New options
464 to verify nonce values and to disable verification. OCSP response
465 printout format cleaned up.
468 *) Add additional OCSP certificate checks. These are those specified
469 in RFC2560. This consists of two separate checks: the CA of the
470 certificate being checked must either be the OCSP signer certificate
471 or the issuer of the OCSP signer certificate. In the latter case the
472 OCSP signer certificate must contain the OCSP signing extended key
473 usage. This check is performed by attempting to match the OCSP
474 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
475 in the OCSP_CERTID structures of the response.
478 *) Initial OCSP certificate verification added to OCSP_basic_verify()
479 and related routines. This uses the standard OpenSSL certificate
480 verify routines to perform initial checks (just CA validity) and
481 to obtain the certificate chain. Then additional checks will be
482 performed on the chain. Currently the root CA is checked to see
483 if it is explicitly trusted for OCSP signing. This is used to set
484 a root CA as a global signing root: that is any certificate that
485 chains to that CA is an acceptable OCSP signing certificate.
488 *) New '-extfile ...' option to 'openssl ca' for reading X.509v3
489 extensions from a separate configuration file.
490 As when reading extensions from the main configuration file,
491 the '-extensions ...' option may be used for specifying the
493 [Massimiliano Pala <madwolf@comune.modena.it>]
495 *) Change PKCS12_key_gen_asc() so it can cope with non null
496 terminated strings whose length is passed in the passlen
497 parameter, for example from PEM callbacks. This was done
498 by adding an extra length parameter to asc2uni().
499 [Steve Henson, reported by <oddissey@samsung.co.kr>]
501 *) New OCSP utility. Allows OCSP requests to be generated or
502 read. The request can be sent to a responder and the output
503 parsed, outputed or printed in text form. Not complete yet:
504 still needs to check the OCSP response validity.
507 *) New subcommands for 'openssl ca':
508 'openssl ca -status <serial>' prints the status of the cert with
509 the given serial number (according to the index file).
510 'openssl ca -updatedb' updates the expiry status of certificates
512 [Massimiliano Pala <madwolf@comune.modena.it>]
514 *) New '-newreq-nodes' command option to CA.pl. This is like
515 '-newreq', but calls 'openssl req' with the '-nodes' option
516 so that the resulting key is not encrypted.
517 [Damien Miller <djm@mindrot.org>]
519 *) New configuration for the GNU Hurd.
520 [Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte]
522 *) Initial code to implement OCSP basic response verify. This
523 is currently incomplete. Currently just finds the signer's
524 certificate and verifies the signature on the response.
527 *) New SSLeay_version code SSLEAY_DIR to determine the compiled-in
528 value of OPENSSLDIR. This is available via the new '-d' option
529 to 'openssl version', and is also included in 'openssl version -a'.
532 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
533 call failed, free the DSA structure.
536 *) Allowing defining memory allocation callbacks that will be given
537 file name and line number information in additional arguments
538 (a const char* and an int). The basic functionality remains, as
539 well as the original possibility to just replace malloc(),
540 realloc() and free() by functions that do not know about these
541 additional arguments. To register and find out the current
542 settings for extended allocation functions, the following
543 functions are provided:
545 CRYPTO_set_mem_ex_functions
546 CRYPTO_set_locked_mem_ex_functions
547 CRYPTO_get_mem_ex_functions
548 CRYPTO_get_locked_mem_ex_functions
550 These work the same way as CRYPTO_set_mem_functions and friends.
551 CRYPTO_get_[locked_]mem_functions now writes 0 where such an
552 extended allocation function is enabled.
553 Similarly, CRYPTO_get_[locked_]mem_ex_functions writes 0 where
554 a conventional allocation function is enabled.
555 [Richard Levitte, Bodo Moeller]
557 *) Fix to uni2asc() to cope with zero length Unicode strings.
558 These are present in some PKCS#12 files.
561 *) Finish off removing the remaining LHASH function pointer casts.
562 There should no longer be any prototype-casting required when using
563 the LHASH abstraction, and any casts that remain are "bugs". See
564 the callback types and macros at the head of lhash.h for details
565 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
568 *) Add automatic query of EGD sockets in RAND_poll() for the unix variant.
569 If an EGD or PRNGD is running and enough entropy is returned, automatic
570 seeding like with /dev/[u]random will be performed.
571 Positions tried are: /etc/entropy, /var/run/egd-pool.
574 *) Change the Unix RAND_poll() variant to be able to poll several
575 random devices, as specified by DEVRANDOM, until a sufficient amount
576 of data has been collected. We spend at most 10 ms on each file
577 (select timeout) and read in non-blocking mode. DEVRANDOM now
578 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
579 (previously it was just the string "/dev/urandom"), so on typical
580 platforms the 10 ms delay will never occur.
581 Also separate out the Unix variant to its own file, rand_unix.c.
582 For VMS, there's a currently-empty rand_vms.c.
585 *) Move OCSP client related routines to ocsp_cl.c. These
586 provide utility functions which an application needing
587 to issue a request to an OCSP responder and analyse the
588 response will typically need: as opposed to those which an
589 OCSP responder itself would need which will be added later.
591 OCSP_request_sign() signs an OCSP request with an API similar
592 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
593 response. OCSP_response_get1_basic() extracts basic response
594 from response. OCSP_resp_find_status(): finds and extracts status
595 information from an OCSP_CERTID structure (which will be created
596 when the request structure is built). These are built from lower
597 level functions which work on OCSP_SINGLERESP structures but
598 wont normally be used unless the application wishes to examine
599 extensions in the OCSP response for example.
601 Replace nonce routines with a pair of functions.
602 OCSP_request_add1_nonce() adds a nonce value and optionally
603 generates a random value. OCSP_check_nonce() checks the
604 validity of the nonce in an OCSP response.
607 *) Change function OCSP_request_add() to OCSP_request_add0_id().
608 This doesn't copy the supplied OCSP_CERTID and avoids the
609 need to free up the newly created id. Change return type
610 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
611 This can then be used to add extensions to the request.
612 Deleted OCSP_request_new(), since most of its functionality
613 is now in OCSP_REQUEST_new() (and the case insensitive name
614 clash) apart from the ability to set the request name which
615 will be added elsewhere.
618 *) Update OCSP API. Remove obsolete extensions argument from
619 various functions. Extensions are now handled using the new
620 OCSP extension code. New simple OCSP HTTP function which
621 can be used to send requests and parse the response.
624 *) Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
625 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
626 uses the special reorder version of SET OF to sort the attributes
627 and reorder them to match the encoded order. This resolves a long
628 standing problem: a verify on a PKCS7 structure just after signing
629 it used to fail because the attribute order did not match the
630 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
631 it uses the received order. This is necessary to tolerate some broken
632 software that does not order SET OF. This is handled by encoding
633 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
634 to produce the required SET OF.
637 *) Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
638 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
639 files to get correct declarations of the ASN.1 item variables.
642 *) Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
643 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
644 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
645 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
646 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
647 ASN1_ITEM and no wrapper functions.
650 *) New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
651 replace the old function pointer based I/O routines. Change most of
652 the *_d2i_bio() and *_d2i_fp() functions to use these.
655 *) Enhance mkdef.pl to be more accepting about spacing in C preprocessor
656 lines, recognice more "algorithms" that can be deselected, and make
657 it complain about algorithm deselection that isn't recognised.
660 *) New ASN1 functions to handle dup, sign, verify, digest, pack and
661 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
662 to use new functions. Add NO_ASN1_OLD which can be set to remove
663 some old style ASN1 functions: this can be used to determine if old
664 code will still work when these eventually go away.
667 *) New extension functions for OCSP structures, these follow the
668 same conventions as certificates and CRLs.
671 *) New function X509V3_add1_i2d(). This automatically encodes and
672 adds an extension. Its behaviour can be customised with various
673 flags to append, replace or delete. Various wrappers added for
674 certifcates and CRLs.
677 *) Fix to avoid calling the underlying ASN1 print routine when
678 an extension cannot be parsed. Correct a typo in the
679 OCSP_SERVICELOC extension. Tidy up print OCSP format.
682 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
683 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
684 when writing a 32767 byte record.
685 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>]
687 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c),
688 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}.
690 (RSA objects have a reference count access to which is protected
691 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
692 so they are meant to be shared between threads.)
693 [Bodo Moeller, Geoff Thorpe; original patch submitted by
694 "Reddie, Steven" <Steven.Reddie@ca.com>]
696 *) Make mkdef.pl parse some of the ASN1 macros and add apropriate
697 entries for variables.
700 *) Fix a deadlock in CRYPTO_mem_leaks().
703 *) Add functionality to apps/openssl.c for detecting locking
704 problems: As the program is single-threaded, all we have
705 to do is register a locking callback using an array for
706 storing which locks are currently held by the program.
709 *) Use a lock around the call to CRYPTO_get_ex_new_index() in
710 SSL_get_ex_data_X509_STORE_idx(), which is used in
711 ssl_verify_cert_chain() and thus can be called at any time
712 during TLS/SSL handshakes so that thread-safety is essential.
713 Unfortunately, the ex_data design is not at all suited
714 for multi-threaded use, so it probably should be abolished.
717 *) Added Broadcom "ubsec" ENGINE to OpenSSL.
718 [Broadcom, tweaked and integrated by Geoff Thorpe]
720 *) Move common extension printing code to new function
721 X509V3_print_extensions(). Reorganise OCSP print routines and
722 implement some needed OCSP ASN1 functions. Add OCSP extensions.
725 *) New function X509_signature_print() to remove duplication in some
729 *) Add a special meaning when SET OF and SEQUENCE OF flags are both
730 set (this was treated exactly the same as SET OF previously). This
731 is used to reorder the STACK representing the structure to match the
732 encoding. This will be used to get round a problem where a PKCS7
733 structure which was signed could not be verified because the STACK
734 order did not reflect the encoded order.
737 *) Reimplement the OCSP ASN1 module using the new code.
740 *) Update the X509V3 code to permit the use of an ASN1_ITEM structure
741 for its ASN1 operations. The old style function pointers still exist
742 for now but they will eventually go away.
745 *) Merge in replacement ASN1 code from the ASN1 branch. This almost
746 completely replaces the old ASN1 functionality with a table driven
747 encoder and decoder which interprets an ASN1_ITEM structure describing
748 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
749 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
750 has also been converted to the new form.
753 *) Change BN_mod_exp_recp so that negative moduli are tolerated
754 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
755 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
759 *) Fix BN_uadd and BN_usub: Always return non-negative results instead
760 of not touching the result's sign bit.
763 *) BN_div bugfix: If the result is 0, the sign (res->neg) must not be
767 *) Changed the LHASH code to use prototypes for callbacks, and created
768 macros to declare and implement thin (optionally static) functions
769 that provide type-safety and avoid function pointer casting for the
770 type-specific callbacks.
773 *) Use better test patterns in bntest.
776 *) Added Kerberos Cipher Suites to be used with TLS, as written in
778 [Veers Staats <staatsvr@asc.hpc.mil>,
779 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte]
781 *) rand_win.c fix for Borland C.
784 *) BN_rshift bugfix for n == 0.
787 *) Reformat the FAQ so the different questions and answers can be divided
788 in sections depending on the subject.
791 *) Have the zlib compression code load ZLIB.DLL dynamically under
795 *) New function BN_mod_sqrt for computing square roots modulo a prime
796 (using the probabilistic Tonelli-Shanks algorithm unless
797 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
798 be handled deterministically).
799 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
801 *) Store verify_result within SSL_SESSION also for client side to
802 avoid potential security hole. (Re-used sessions on the client side
803 always resulted in verify_result==X509_V_OK, not using the original
804 result of the server certificate verification.)
807 *) Make BN_mod_inverse faster by explicitly handling small quotients
808 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
809 512 bits], about 30% for larger ones [1024 or 2048 bits].)
812 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type
813 SSL3_RT_APPLICATION_DATA, return 0.
814 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
818 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
819 releases, have been re-implemented by renaming the previous
820 implementations of ssl2_read and ssl3_read to ssl2_read_internal
821 and ssl3_read_internal, respectively, and adding 'peek' parameters
822 to them. The new ssl[23]_{read,peek} functions are calls to
823 ssl[23]_read_internal with the 'peek' flag set appropriately.
824 A 'peek' parameter has also been added to ssl3_read_bytes, which
825 does the actual work for ssl3_read_internal.
828 *) New function BN_kronecker.
831 *) Fix BN_gcd so that it works on negative inputs; the result is
832 positive unless both parameters are zero.
833 Previously something reasonably close to an infinite loop was
834 possible because numbers could be growing instead of shrinking
835 in the implementation of Euclid's algorithm.
838 *) Fix BN_is_word() and BN_is_one() macros to take into account the
839 sign of the number in question.
841 Fix BN_is_word(a,w) to work correctly for w == 0.
843 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
844 because its test if the absolute value of 'a' equals 'w'.
845 Note that BN_abs_is_word does *not* handle w == 0 reliably;
846 it exists mostly for use in the implementations of BN_is_zero(),
847 BN_is_one(), and BN_is_word().
850 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
851 the method-specific "init()" handler. Also clean up ex_data after
852 calling the method-specific "finish()" handler. Previously, this was
853 happening the other way round.
856 *) New function BN_swap.
859 *) Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
860 the exponentiation functions are more likely to produce reasonable
861 results on negative inputs.
864 *) Change BN_mod_mul so that the result is always non-negative.
865 Previously, it could be negative if one of the factors was negative;
866 I don't think anyone really wanted that behaviour.
869 *) Move BN_mod_... functions into new file crypto/bn/bn_mod.c
870 (except for exponentiation, which stays in crypto/bn/bn_exp.c,
871 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c)
872 and add new functions:
885 These functions always generate non-negative results.
887 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r
888 such that |m| < r < 0, BN_nnmod will output rem + |m| instead).
890 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as
891 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b]
893 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
895 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
896 was actually never needed) and in BN_mul(). The removal in BN_mul()
897 required a small change in bn_mul_part_recursive() and the addition
898 of the functions bn_cmp_part_words(), bn_sub_part_words() and
899 bn_add_part_words(), which do the same thing as bn_cmp_words(),
900 bn_sub_words() and bn_add_words() except they take arrays with
904 *) In 'openssl passwd', verify passwords read from the terminal
905 unless the '-salt' option is used (which usually means that
906 verification would just waste user's time since the resulting
907 hash is going to be compared with some given password hash)
908 or the new '-noverify' option is used.
910 This is an incompatible change, but it does not affect
911 non-interactive use of 'openssl passwd' (passwords on the command
912 line, '-stdin' option, '-in ...' option) and thus should not
916 *) Remove all references to RSAref, since there's no more need for it.
919 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
920 The previous value, 12, was not always sufficient for BN_mod_exp().
923 *) Make DSO load along a path given through an environment variable
924 (SHLIB_PATH) with shl_load().
927 *) Constify the ENGINE code as a result of BIGNUM constification.
928 Also constify the RSA code and most things related to it. In a
929 few places, most notable in the depth of the ASN.1 code, ugly
930 casts back to non-const were required (to be solved at a later
934 *) Make it so the openssl application has all engines loaded by default.
937 *) Constify the BIGNUM routines a little more.
940 *) Make sure that shared libraries get the internal name engine with
941 the full version number and not just 0. This should mark the
942 shared libraries as not backward compatible. Of course, this should
943 be changed again when we can guarantee backward binary compatibility.
946 *) Add the following functions:
952 ENGINE_load_builtin_engines()
954 That way, an application can itself choose if external engines that
955 are built-in in OpenSSL shall ever be used or not. The benefit is
956 that applications won't have to be linked with libdl or other dso
957 libraries unless it's really needed.
959 Changed 'openssl engine' to load all engines on demand.
960 Changed the engine header files to avoid the duplication of some
961 declarations (they differed!).
964 *) 'openssl engine' can now list capabilities.
967 *) Better error reporting in 'openssl engine'.
970 *) Never call load_dh_param(NULL) in s_server.
973 *) Add engine application. It can currently list engines by name and
974 identity, and test if they are actually available.
977 *) Add support for shared libraries under Irix.
978 [Albert Chin-A-Young <china@thewrittenword.com>]
980 *) Improve RPM specification file by forcing symbolic linking and making
981 sure the installed documentation is also owned by root.root.
982 [Damien Miller <djm@mindrot.org>]
984 *) Add configuration option to build on Linux on both big-endian and
986 [Ralf Baechle <ralf@uni-koblenz.de>]
988 *) Give the OpenSSL applications more possibilities to make use of
989 keys (public as well as private) handled by engines.
992 *) Add OCSP code that comes from CertCo.
995 *) Add VMS support for the Rijndael code.
998 *) Added untested support for Nuron crypto accelerator.
1001 *) Add support for external cryptographic devices. This code was
1002 previously distributed separately as the "engine" branch.
1003 [Geoff Thorpe, Richard Levitte]
1005 *) Rework the filename-translation in the DSO code. It is now possible to
1006 have far greater control over how a "name" is turned into a filename
1007 depending on the operating environment and any oddities about the
1008 different shared library filenames on each system.
1011 *) Support threads on FreeBSD-elf in Configure.
1014 *) Add the possibility to create shared libraries on HP-UX
1017 *) Fix for SHA1 assembly problem with MASM: it produces
1018 warnings about corrupt line number information when assembling
1019 with debugging information. This is caused by the overlapping
1021 [Bernd Matthes <mainbug@celocom.de>, Steve Henson]
1024 NCONF_get_number() has no error checking at all. As a replacement,
1025 NCONF_get_number_e() is defined (_e for "error checking") and is
1026 promoted strongly. The old NCONF_get_number is kept around for
1027 binary backward compatibility.
1028 Make it possible for methods to load from something other than a BIO,
1029 by providing a function pointer that is given a name instead of a BIO.
1030 For example, this could be used to load configuration data from an
1034 *) Fix typo in get_cert_by_subject() in by_dir.c
1035 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>]
1037 *) Rework the system to generate shared libraries:
1039 - Make note of the expected extension for the shared libraries and
1040 if there is a need for symbolic links from for example libcrypto.so.0
1041 to libcrypto.so.0.9.7. There is extended info in Configure for
1044 - Make as few rebuilds of the shared libraries as possible.
1046 - Still avoid linking the OpenSSL programs with the shared libraries.
1048 - When installing, install the shared libraries separately from the
1052 *) Fix for non blocking accept BIOs. Added new I/O special reason
1053 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
1054 with non blocking I/O was not possible because no retry code was
1055 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
1059 *) Added the beginnings of Rijndael support.
1062 *) Fix for bug in DirectoryString mask setting. Add support for
1063 X509_NAME_print_ex() in 'req' and X509_print_ex() function
1064 to allow certificate printing to more controllable, additional
1065 'certopt' option to 'x509' to allow new printing options to be
1069 *) Clean old EAY MD5 hack from e_os.h.
1072 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument.
1074 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
1075 and not in SSL_clear because the latter is also used by the
1076 accept/connect functions; previously, the settings made by
1077 SSL_set_read_ahead would be lost during the handshake.
1078 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>]
1080 *) Correct util/mkdef.pl to be selective about disabled algorithms.
1081 Previously, it would create entries for disableed algorithms no
1085 Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
1087 *) In ssl23_get_client_hello, generate an error message when faced
1088 with an initial SSL 3.0/TLS record that is too small to contain the
1089 first two bytes of the ClientHello message, i.e. client_version.
1090 (Note that this is a pathologic case that probably has never happened
1091 in real life.) The previous approach was to use the version number
1092 from the record header as a substitute; but our protocol choice
1093 should not depend on that one because it is not authenticated
1094 by the Finished messages.
1097 *) More robust randomness gathering functions for Windows.
1098 [Jeffrey Altman <jaltman@columbia.edu>]
1100 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
1101 not set then we don't setup the error code for issuer check errors
1102 to avoid possibly overwriting other errors which the callback does
1103 handle. If an application does set the flag then we assume it knows
1104 what it is doing and can handle the new informational codes
1108 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
1109 a general "ANY" type, as such it should be able to decode anything
1110 including tagged types. However it didn't check the class so it would
1111 wrongly interpret tagged types in the same way as their universal
1112 counterpart and unknown types were just rejected. Changed so that the
1113 tagged and unknown types are handled in the same way as a SEQUENCE:
1114 that is the encoding is stored intact. There is also a new type
1115 "V_ASN1_OTHER" which is used when the class is not universal, in this
1116 case we have no idea what the actual type is so we just lump them all
1120 *) On VMS, stdout may very well lead to a file that is written to
1121 in a record-oriented fashion. That means that every write() will
1122 write a separate record, which will be read separately by the
1123 programs trying to read from it. This can be very confusing.
1125 The solution is to put a BIO filter in the way that will buffer
1126 text until a linefeed is reached, and then write everything a
1127 line at a time, so every record written will be an actual line,
1128 not chunks of lines and not (usually doesn't happen, but I've
1129 seen it once) several lines in one record. BIO_f_linebuffer() is
1132 Currently, it's a VMS-only method, because that's where it has
1133 been tested well enough.
1136 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
1137 it can return incorrect results.
1138 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
1139 but it was in 0.9.6-beta[12].)
1142 *) Disable the check for content being present when verifying detached
1143 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
1144 include zero length content when signing messages.
1147 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
1148 BIO_ctrl (for BIO pairs).
1151 *) Add DSO method for VMS.
1154 *) Bug fix: Montgomery multiplication could produce results with the
1158 *) Add RPM specification openssl.spec and modify it to build three
1159 packages. The default package contains applications, application
1160 documentation and run-time libraries. The devel package contains
1161 include files, static libraries and function documentation. The
1162 doc package contains the contents of the doc directory. The original
1163 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
1166 *) Add a large number of documentation files for many SSL routines.
1167 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
1169 *) Add a configuration entry for Sony News 4.
1170 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>]
1172 *) Don't set the two most significant bits to one when generating a
1173 random number < q in the DSA library.
1176 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
1177 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
1178 the underlying transport is blocking) if a handshake took place.
1179 (The default behaviour is needed by applications such as s_client
1180 and s_server that use select() to determine when to use SSL_read;
1181 but for applications that know in advance when to expect data, it
1182 just makes things more complicated.)
1185 *) Add RAND_egd_bytes(), which gives control over the number of bytes read
1189 *) Add a few more EBCDIC conditionals that make `req' and `x509'
1190 work better on such systems.
1191 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1193 *) Add two demo programs for PKCS12_parse() and PKCS12_create().
1194 Update PKCS12_parse() so it copies the friendlyName and the
1195 keyid to the certificates aux info.
1198 *) Fix bug in PKCS7_verify() which caused an infinite loop
1199 if there was more than one signature.
1200 [Sven Uszpelkat <su@celocom.de>]
1202 *) Major change in util/mkdef.pl to include extra information
1203 about each symbol, as well as presentig variables as well
1204 as functions. This change means that there's n more need
1205 to rebuild the .num files when some algorithms are excluded.
1208 *) Allow the verify time to be set by an application,
1209 rather than always using the current time.
1212 *) Phase 2 verify code reorganisation. The certificate
1213 verify code now looks up an issuer certificate by a
1214 number of criteria: subject name, authority key id
1215 and key usage. It also verifies self signed certificates
1216 by the same criteria. The main comparison function is
1217 X509_check_issued() which performs these checks.
1219 Lot of changes were necessary in order to support this
1220 without completely rewriting the lookup code.
1222 Authority and subject key identifier are now cached.
1224 The LHASH 'certs' is X509_STORE has now been replaced
1225 by a STACK_OF(X509_OBJECT). This is mainly because an
1226 LHASH can't store or retrieve multiple objects with
1227 the same hash value.
1229 As a result various functions (which were all internal
1230 use only) have changed to handle the new X509_STORE
1231 structure. This will break anything that messed round
1232 with X509_STORE internally.
1234 The functions X509_STORE_add_cert() now checks for an
1235 exact match, rather than just subject name.
1237 The X509_STORE API doesn't directly support the retrieval
1238 of multiple certificates matching a given criteria, however
1239 this can be worked round by performing a lookup first
1240 (which will fill the cache with candidate certificates)
1241 and then examining the cache for matches. This is probably
1242 the best we can do without throwing out X509_LOOKUP
1243 entirely (maybe later...).
1245 The X509_VERIFY_CTX structure has been enhanced considerably.
1247 All certificate lookup operations now go via a get_issuer()
1248 callback. Although this currently uses an X509_STORE it
1249 can be replaced by custom lookups. This is a simple way
1250 to bypass the X509_STORE hackery necessary to make this
1251 work and makes it possible to use more efficient techniques
1252 in future. A very simple version which uses a simple
1253 STACK for its trusted certificate store is also provided
1254 using X509_STORE_CTX_trusted_stack().
1256 The verify_cb() and verify() callbacks now have equivalents
1257 in the X509_STORE_CTX structure.
1259 X509_STORE_CTX also has a 'flags' field which can be used
1260 to customise the verify behaviour.
1263 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
1264 excludes S/MIME capabilities.
1267 *) When a certificate request is read in keep a copy of the
1268 original encoding of the signed data and use it when outputing
1269 again. Signatures then use the original encoding rather than
1270 a decoded, encoded version which may cause problems if the
1271 request is improperly encoded.
1274 *) For consistency with other BIO_puts implementations, call
1275 buffer_write(b, ...) directly in buffer_puts instead of calling
1278 In BIO_puts, increment b->num_write as in BIO_write.
1279 [Peter.Sylvester@EdelWeb.fr]
1281 *) Fix BN_mul_word for the case where the word is 0. (We have to use
1282 BN_zero, we may not return a BIGNUM with an array consisting of
1286 *) Avoid calling abort() from within the library when problems are
1287 detected, except if preprocessor symbols have been defined
1288 (such as REF_CHECK, BN_DEBUG etc.).
1291 *) New openssl application 'rsautl'. This utility can be
1292 used for low level RSA operations. DER public key
1293 BIO/fp routines also added.
1296 *) New Configure entry and patches for compiling on QNX 4.
1297 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
1299 *) A demo state-machine implementation was sponsored by
1300 Nuron (http://www.nuron.com/) and is now available in
1301 demos/state_machine.
1304 *) New options added to the 'dgst' utility for signature
1305 generation and verification.
1308 *) Unrecognized PKCS#7 content types are now handled via a
1309 catch all ASN1_TYPE structure. This allows unsupported
1310 types to be stored as a "blob" and an application can
1311 encode and decode it manually.
1314 *) Fix various signed/unsigned issues to make a_strex.c
1316 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
1318 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
1319 length if passed a buffer. ASN1_INTEGER_to_BN failed
1320 if passed a NULL BN and its argument was negative.
1321 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
1323 *) Modification to PKCS#7 encoding routines to output definite
1324 length encoding. Since currently the whole structures are in
1325 memory there's not real point in using indefinite length
1326 constructed encoding. However if OpenSSL is compiled with
1327 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
1330 *) Added BIO_vprintf() and BIO_vsnprintf().
1333 *) Added more prefixes to parse for in the the strings written
1334 through a logging bio, to cover all the levels that are available
1335 through syslog. The prefixes are now:
1337 PANIC, EMERG, EMR => LOG_EMERG
1338 ALERT, ALR => LOG_ALERT
1339 CRIT, CRI => LOG_CRIT
1340 ERROR, ERR => LOG_ERR
1341 WARNING, WARN, WAR => LOG_WARNING
1342 NOTICE, NOTE, NOT => LOG_NOTICE
1343 INFO, INF => LOG_INFO
1344 DEBUG, DBG => LOG_DEBUG
1346 and as before, if none of those prefixes are present at the
1347 beginning of the string, LOG_ERR is chosen.
1349 On Win32, the LOG_* levels are mapped according to this:
1351 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
1352 LOG_WARNING => EVENTLOG_WARNING_TYPE
1353 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
1357 *) Made it possible to reconfigure with just the configuration
1358 argument "reconf" or "reconfigure". The command line arguments
1359 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
1360 and are retrieved from there when reconfiguring.
1364 [Assar Westerlund <assar@sics.se>, Richard Levitte]
1366 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
1369 *) The obj_dat.pl script was messing up the sorting of object
1370 names. The reason was that it compared the quoted version
1371 of strings as a result "OCSP" > "OCSP Signing" because
1372 " > SPACE. Changed script to store unquoted versions of
1373 names and add quotes on output. It was also omitting some
1374 names from the lookup table if they were given a default
1375 value (that is if SN is missing it is given the same
1376 value as LN and vice versa), these are now added on the
1377 grounds that if an object has a name we should be able to
1378 look it up. Finally added warning output when duplicate
1379 short or long names are found.
1382 *) Changes needed for Tandem NSK.
1383 [Scott Uroff <scott@xypro.com>]
1385 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
1386 RSA_padding_check_SSLv23(), special padding was never detected
1387 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
1388 version rollback attacks was not effective.
1390 In s23_clnt.c, don't use special rollback-attack detection padding
1391 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
1392 client; similarly, in s23_srvr.c, don't do the rollback check if
1393 SSL 2.0 is the only protocol enabled in the server.
1396 *) Make it possible to get hexdumps of unprintable data with 'openssl
1397 asn1parse'. By implication, the functions ASN1_parse_dump() and
1398 BIO_dump_indent() are added.
1401 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
1402 these print out strings and name structures based on various
1403 flags including RFC2253 support and proper handling of
1404 multibyte characters. Added options to the 'x509' utility
1405 to allow the various flags to be set.
1408 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
1409 Also change the functions X509_cmp_current_time() and
1410 X509_gmtime_adj() work with an ASN1_TIME structure,
1411 this will enable certificates using GeneralizedTime in validity
1412 dates to be checked.
1415 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
1416 negative public key encodings) on by default,
1417 NO_NEG_PUBKEY_BUG can be set to disable it.
1420 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
1421 content octets. An i2c_ASN1_OBJECT is unnecessary because
1422 the encoding can be trivially obtained from the structure.
1425 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
1426 not read locks (CRYPTO_r_[un]lock).
1429 *) A first attempt at creating official support for shared
1430 libraries through configuration. I've kept it so the
1431 default is static libraries only, and the OpenSSL programs
1432 are always statically linked for now, but there are
1433 preparations for dynamic linking in place.
1434 This has been tested on Linux and True64.
1437 *) Randomness polling function for Win9x, as described in:
1438 Peter Gutmann, Software Generation of Practically Strong
1442 *) Fix so PRNG is seeded in req if using an already existing
1446 *) New options to smime application. -inform and -outform
1447 allow alternative formats for the S/MIME message including
1448 PEM and DER. The -content option allows the content to be
1449 specified separately. This should allow things like Netscape
1450 form signing output easier to verify.
1453 *) Fix the ASN1 encoding of tags using the 'long form'.
1456 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
1457 STRING types. These convert content octets to and from the
1458 underlying type. The actual tag and length octets are
1459 already assumed to have been read in and checked. These
1460 are needed because all other string types have virtually
1461 identical handling apart from the tag. By having versions
1462 of the ASN1 functions that just operate on content octets
1463 IMPLICIT tagging can be handled properly. It also allows
1464 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
1465 and ASN1_INTEGER are identical apart from the tag.
1468 *) Change the handling of OID objects as follows:
1470 - New object identifiers are inserted in objects.txt, following
1471 the syntax given in objects.README.
1472 - objects.pl is used to process obj_mac.num and create a new
1474 - obj_dat.pl is used to create a new obj_dat.h, using the data in
1477 This is currently kind of a hack, and the perl code in objects.pl
1478 isn't very elegant, but it works as I intended. The simplest way
1479 to check that it worked correctly is to look in obj_dat.h and
1480 check the array nid_objs and make sure the objects haven't moved
1481 around (this is important!). Additions are OK, as well as
1482 consistent name changes.
1485 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
1488 *) Addition of the command line parameter '-rand file' to 'openssl req'.
1489 The given file adds to whatever has already been seeded into the
1490 random pool through the RANDFILE configuration file option or
1491 environment variable, or the default random state file.
1494 *) mkstack.pl now sorts each macro group into lexical order.
1495 Previously the output order depended on the order the files
1496 appeared in the directory, resulting in needless rewriting
1500 *) Patches to make OpenSSL compile under Win32 again. Mostly
1501 work arounds for the VC++ problem that it treats func() as
1502 func(void). Also stripped out the parts of mkdef.pl that
1503 added extra typesafe functions: these no longer exist.
1506 *) Reorganisation of the stack code. The macros are now all
1507 collected in safestack.h . Each macro is defined in terms of
1508 a "stack macro" of the form SKM_<name>(type, a, b). The
1509 DEBUG_SAFESTACK is now handled in terms of function casts,
1510 this has the advantage of retaining type safety without the
1511 use of additional functions. If DEBUG_SAFESTACK is not defined
1512 then the non typesafe macros are used instead. Also modified the
1513 mkstack.pl script to handle the new form. Needs testing to see
1514 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
1515 the default if no major problems. Similar behaviour for ASN1_SET_OF
1516 and PKCS12_STACK_OF.
1519 *) When some versions of IIS use the 'NET' form of private key the
1520 key derivation algorithm is different. Normally MD5(password) is
1521 used as a 128 bit RC4 key. In the modified case
1522 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
1523 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
1524 as the old Netscape_RSA functions except they have an additional
1525 'sgckey' parameter which uses the modified algorithm. Also added
1526 an -sgckey command line option to the rsa utility. Thanks to
1527 Adrian Peck <bertie@ncipher.com> for posting details of the modified
1528 algorithm to openssl-dev.
1531 *) The evp_local.h macros were using 'c.##kname' which resulted in
1532 invalid expansion on some systems (SCO 5.0.5 for example).
1533 Corrected to 'c.kname'.
1534 [Phillip Porch <root@theporch.com>]
1536 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
1537 a STACK of email addresses from a certificate or request, these look
1538 in the subject name and the subject alternative name extensions and
1539 omit any duplicate addresses.
1542 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
1543 This makes DSA verification about 2 % faster.
1546 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
1547 (meaning that now 2^5 values will be precomputed, which is only 4 KB
1548 plus overhead for 1024 bit moduli).
1549 This makes exponentiations about 0.5 % faster for 1024 bit
1550 exponents (as measured by "openssl speed rsa2048").
1553 *) Rename memory handling macros to avoid conflicts with other
1555 Malloc => OPENSSL_malloc
1556 Malloc_locked => OPENSSL_malloc_locked
1557 Realloc => OPENSSL_realloc
1558 Free => OPENSSL_free
1561 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
1562 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
1565 *) CygWin32 support.
1566 [John Jarvie <jjarvie@newsguy.com>]
1568 *) The type-safe stack code has been rejigged. It is now only compiled
1569 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
1570 by default all type-specific stack functions are "#define"d back to
1571 standard stack functions. This results in more streamlined output
1572 but retains the type-safety checking possibilities of the original
1576 *) The STACK code has been cleaned up, and certain type declarations
1577 that didn't make a lot of sense have been brought in line. This has
1578 also involved a cleanup of sorts in safestack.h to more correctly
1579 map type-safe stack functions onto their plain stack counterparts.
1580 This work has also resulted in a variety of "const"ifications of
1581 lots of the code, especially "_cmp" operations which should normally
1582 be prototyped with "const" parameters anyway.
1585 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
1586 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
1587 (The PRNG state consists of two parts, the large pool 'state' and 'md',
1588 where all of 'md' is used each time the PRNG is used, but 'state'
1589 is used only indexed by a cyclic counter. As entropy may not be
1590 well distributed from the beginning, 'md' is important as a
1591 chaining variable. However, the output function chains only half
1592 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
1593 all of 'md', and seeding with STATE_SIZE dummy bytes will result
1594 in all of 'state' being rewritten, with the new values depending
1595 on virtually all of 'md'. This overcomes the 80 bit limitation.)
1598 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
1599 the handshake is continued after ssl_verify_cert_chain();
1600 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
1601 can lead to 'unexplainable' connection aborts later.
1602 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
1604 *) Major EVP API cipher revision.
1605 Add hooks for extra EVP features. This allows various cipher
1606 parameters to be set in the EVP interface. Support added for variable
1607 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
1608 setting of RC2 and RC5 parameters.
1610 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
1613 Remove lots of duplicated code from the EVP library. For example *every*
1614 cipher init() function handles the 'iv' in the same way according to the
1615 cipher mode. They also all do nothing if the 'key' parameter is NULL and
1616 for CFB and OFB modes they zero ctx->num.
1618 New functionality allows removal of S/MIME code RC2 hack.
1620 Most of the routines have the same form and so can be declared in terms
1623 By shifting this to the top level EVP_CipherInit() it can be removed from
1624 all individual ciphers. If the cipher wants to handle IVs or keys
1625 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
1628 Change lots of functions like EVP_EncryptUpdate() to now return a
1629 value: although software versions of the algorithms cannot fail
1630 any installed hardware versions can.
1633 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
1634 this option is set, tolerate broken clients that send the negotiated
1635 protocol version number instead of the requested protocol version
1639 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
1640 i.e. non-zero for export ciphersuites, zero otherwise.
1641 Previous versions had this flag inverted, inconsistent with
1642 rsa_tmp_cb (..._TMP_RSA_CB).
1643 [Bodo Moeller; problem reported by Amit Chopra]
1645 *) Add missing DSA library text string. Work around for some IIS
1646 key files with invalid SEQUENCE encoding.
1649 *) Add a document (doc/standards.txt) that list all kinds of standards
1650 and so on that are implemented in OpenSSL.
1653 *) Enhance c_rehash script. Old version would mishandle certificates
1654 with the same subject name hash and wouldn't handle CRLs at all.
1655 Added -fingerprint option to crl utility, to support new c_rehash
1659 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
1662 *) Fix for SSL server purpose checking. Server checking was
1663 rejecting certificates which had extended key usage present
1664 but no ssl client purpose.
1665 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
1667 *) Make PKCS#12 code work with no password. The PKCS#12 spec
1668 is a little unclear about how a blank password is handled.
1669 Since the password in encoded as a BMPString with terminating
1670 double NULL a zero length password would end up as just the
1671 double NULL. However no password at all is different and is
1672 handled differently in the PKCS#12 key generation code. NS
1673 treats a blank password as zero length. MSIE treats it as no
1674 password on export: but it will try both on import. We now do
1675 the same: PKCS12_parse() tries zero length and no password if
1676 the password is set to "" or NULL (NULL is now a valid password:
1677 it wasn't before) as does the pkcs12 application.
1680 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
1681 perror when PEM_read_bio_X509_REQ fails, the error message must
1682 be obtained from the error queue.
1685 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
1686 it in ERR_remove_state if appropriate, and change ERR_get_state
1687 accordingly to avoid race conditions (this is necessary because
1688 thread_hash is no longer constant once set).
1691 *) Bugfix for linux-elf makefile.one.
1694 *) RSA_get_default_method() will now cause a default
1695 RSA_METHOD to be chosen if one doesn't exist already.
1696 Previously this was only set during a call to RSA_new()
1697 or RSA_new_method(NULL) meaning it was possible for
1698 RSA_get_default_method() to return NULL.
1701 *) Added native name translation to the existing DSO code
1702 that will convert (if the flag to do so is set) filenames
1703 that are sufficiently small and have no path information
1704 into a canonical native form. Eg. "blah" converted to
1705 "libblah.so" or "blah.dll" etc.
1708 *) New function ERR_error_string_n(e, buf, len) which is like
1709 ERR_error_string(e, buf), but writes at most 'len' bytes
1710 including the 0 terminator. For ERR_error_string_n, 'buf'
1712 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
1714 *) CONF library reworked to become more general. A new CONF
1715 configuration file reader "class" is implemented as well as a
1716 new functions (NCONF_*, for "New CONF") to handle it. The now
1717 old CONF_* functions are still there, but are reimplemented to
1718 work in terms of the new functions. Also, a set of functions
1719 to handle the internal storage of the configuration data is
1720 provided to make it easier to write new configuration file
1721 reader "classes" (I can definitely see something reading a
1722 configuration file in XML format, for example), called _CONF_*,
1723 or "the configuration storage API"...
1725 The new configuration file reading functions are:
1727 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
1728 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
1730 NCONF_default, NCONF_WIN32
1732 NCONF_dump_fp, NCONF_dump_bio
1734 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
1735 NCONF_new creates a new CONF object. This works in the same way
1736 as other interfaces in OpenSSL, like the BIO interface.
1737 NCONF_dump_* dump the internal storage of the configuration file,
1738 which is useful for debugging. All other functions take the same
1739 arguments as the old CONF_* functions wth the exception of the
1740 first that must be a `CONF *' instead of a `LHASH *'.
1742 To make it easer to use the new classes with the old CONF_* functions,
1743 the function CONF_set_default_method is provided.
1746 *) Add '-tls1' option to 'openssl ciphers', which was already
1747 mentioned in the documentation but had not been implemented.
1748 (This option is not yet really useful because even the additional
1749 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
1752 *) Initial DSO code added into libcrypto for letting OpenSSL (and
1753 OpenSSL-based applications) load shared libraries and bind to
1754 them in a portable way.
1755 [Geoff Thorpe, with contributions from Richard Levitte]
1757 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
1759 *) Make sure _lrotl and _lrotr are only used with MSVC.
1761 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
1762 (the default implementation of RAND_status).
1764 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
1765 to '-clrext' (= clear extensions), as intended and documented.
1766 [Bodo Moeller; inconsistency pointed out by Michael Attili
1769 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
1770 was larger than the MD block size.
1771 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
1773 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
1774 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
1775 using the passed key: if the passed key was a private key the result
1776 of X509_print(), for example, would be to print out all the private key
1780 *) des_quad_cksum() byte order bug fix.
1781 [Ulf Möller, using the problem description in krb4-0.9.7, where
1782 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
1784 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
1786 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
1788 *) For easily testing in shell scripts whether some command
1789 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
1790 returns with exit code 0 iff no command of the given name is available.
1791 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
1792 the output goes to stdout and nothing is printed to stderr.
1793 Additional arguments are always ignored.
1795 Since for each cipher there is a command of the same name,
1796 the 'no-cipher' compilation switches can be tested this way.
1798 ('openssl no-XXX' is not able to detect pseudo-commands such
1799 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
1802 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
1805 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
1806 is set; it will be thrown away anyway because each handshake creates
1808 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
1809 to parameters -- in previous versions (since OpenSSL 0.9.3) the
1810 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
1811 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
1814 *) New s_client option -ign_eof: EOF at stdin is ignored, and
1815 'Q' and 'R' lose their special meanings (quit/renegotiate).
1816 This is part of what -quiet does; unlike -quiet, -ign_eof
1817 does not suppress any output.
1820 *) Add compatibility options to the purpose and trust code. The
1821 purpose X509_PURPOSE_ANY is "any purpose" which automatically
1822 accepts a certificate or CA, this was the previous behaviour,
1823 with all the associated security issues.
1825 X509_TRUST_COMPAT is the old trust behaviour: only and
1826 automatically trust self signed roots in certificate store. A
1827 new trust setting X509_TRUST_DEFAULT is used to specify that
1828 a purpose has no associated trust setting and it should instead
1829 use the value in the default purpose.
1832 *) Fix the PKCS#8 DSA private key code so it decodes keys again
1833 and fix a memory leak.
1836 *) In util/mkerr.pl (which implements 'make errors'), preserve
1837 reason strings from the previous version of the .c file, as
1838 the default to have only downcase letters (and digits) in
1839 automatically generated reasons codes is not always appropriate.
1842 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
1843 using strerror. Previously, ERR_reason_error_string() returned
1844 library names as reason strings for SYSerr; but SYSerr is a special
1845 case where small numbers are errno values, not library numbers.
1848 *) Add '-dsaparam' option to 'openssl dhparam' application. This
1849 converts DSA parameters into DH parameters. (When creating parameters,
1850 DSA_generate_parameters is used.)
1853 *) Include 'length' (recommended exponent length) in C code generated
1854 by 'openssl dhparam -C'.
1857 *) The second argument to set_label in perlasm was already being used
1858 so couldn't be used as a "file scope" flag. Moved to third argument
1862 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
1863 instead of RAND_bytes for encryption IVs and salts.
1866 *) Include RAND_status() into RAND_METHOD instead of implementing
1867 it only for md_rand.c Otherwise replacing the PRNG by calling
1868 RAND_set_rand_method would be impossible.
1871 *) Don't let DSA_generate_key() enter an infinite loop if the random
1872 number generation fails.
1875 *) New 'rand' application for creating pseudo-random output.
1878 *) Added configuration support for Linux/IA64
1879 [Rolf Haberrecker <rolf@suse.de>]
1881 *) Assembler module support for Mingw32.
1884 *) Shared library support for HPUX (in shlib/).
1885 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
1887 *) Shared library support for Solaris gcc.
1888 [Lutz Behnke <behnke@trustcenter.de>]
1890 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
1892 *) PKCS7_encrypt() was adding text MIME headers twice because they
1893 were added manually and by SMIME_crlf_copy().
1896 *) In bntest.c don't call BN_rand with zero bits argument.
1897 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
1899 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
1900 case was implemented. This caused BN_div_recp() to fail occasionally.
1903 *) Add an optional second argument to the set_label() in the perl
1904 assembly language builder. If this argument exists and is set
1905 to 1 it signals that the assembler should use a symbol whose
1906 scope is the entire file, not just the current function. This
1907 is needed with MASM which uses the format label:: for this scope.
1908 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
1910 *) Change the ASN1 types so they are typedefs by default. Before
1911 almost all types were #define'd to ASN1_STRING which was causing
1912 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
1916 *) Change names of new functions to the new get1/get0 naming
1917 convention: After 'get1', the caller owns a reference count
1918 and has to call ..._free; 'get0' returns a pointer to some
1919 data structure without incrementing reference counters.
1920 (Some of the existing 'get' functions increment a reference
1921 counter, some don't.)
1922 Similarly, 'set1' and 'add1' functions increase reference
1923 counters or duplicate objects.
1926 *) Allow for the possibility of temp RSA key generation failure:
1927 the code used to assume it always worked and crashed on failure.
1930 *) Fix potential buffer overrun problem in BIO_printf().
1931 [Ulf Möller, using public domain code by Patrick Powell; problem
1932 pointed out by David Sacerdote <das33@cornell.edu>]
1934 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
1935 RAND_egd() and RAND_status(). In the command line application,
1936 the EGD socket can be specified like a seed file using RANDFILE
1940 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
1941 Some CAs (e.g. Verisign) distribute certificates in this form.
1944 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
1945 list to exclude them. This means that no special compilation option
1946 is needed to use anonymous DH: it just needs to be included in the
1950 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
1951 EVP_MD_type. The old functionality is available in a new macro called
1952 EVP_MD_md(). Change code that uses it and update docs.
1955 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
1956 where the 'void *' argument is replaced by a function pointer argument.
1957 Previously 'void *' was abused to point to functions, which works on
1958 many platforms, but is not correct. As these functions are usually
1959 called by macros defined in OpenSSL header files, most source code
1960 should work without changes.
1963 *) <openssl/opensslconf.h> (which is created by Configure) now contains
1964 sections with information on -D... compiler switches used for
1965 compiling the library so that applications can see them. To enable
1966 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
1967 must be defined. E.g.,
1968 #define OPENSSL_ALGORITHM_DEFINES
1969 #include <openssl/opensslconf.h>
1970 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
1971 [Richard Levitte, Ulf and Bodo Möller]
1973 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
1977 *) Change the 'other' type in certificate aux info to a STACK_OF
1978 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
1979 the required ASN1 format: arbitrary types determined by an OID.
1982 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
1983 argument to 'req'. This is not because the function is newer or
1984 better than others it just uses the work 'NEW' in the certificate
1985 request header lines. Some software needs this.
1988 *) Reorganise password command line arguments: now passwords can be
1989 obtained from various sources. Delete the PEM_cb function and make
1990 it the default behaviour: i.e. if the callback is NULL and the
1991 usrdata argument is not NULL interpret it as a null terminated pass
1992 phrase. If usrdata and the callback are NULL then the pass phrase
1993 is prompted for as usual.
1996 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
1997 the support is automatically enabled. The resulting binaries will
1998 autodetect the card and use it if present.
1999 [Ben Laurie and Compaq Inc.]
2001 *) Work around for Netscape hang bug. This sends certificate request
2002 and server done in one record. Since this is perfectly legal in the
2003 SSL/TLS protocol it isn't a "bug" option and is on by default. See
2004 the bugs/SSLv3 entry for more info.
2007 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
2010 *) Add -rand argument to smime and pkcs12 applications and read/write
2014 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
2017 *) Add command line password options to the remaining applications.
2020 *) Bug fix for BN_div_recp() for numerators with an even number of
2024 *) More tests in bntest.c, and changed test_bn output.
2027 *) ./config recognizes MacOS X now.
2030 *) Bug fix for BN_div() when the first words of num and divsor are
2031 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
2034 *) Add support for various broken PKCS#8 formats, and command line
2035 options to produce them.
2038 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
2039 get temporary BIGNUMs from a BN_CTX.
2042 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
2046 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
2047 include a #define from the old name to the new. The original intent
2048 was that statically linked binaries could for example just call
2049 SSLeay_add_all_ciphers() to just add ciphers to the table and not
2050 link with digests. This never worked becayse SSLeay_add_all_digests()
2051 and SSLeay_add_all_ciphers() were in the same source file so calling
2052 one would link with the other. They are now in separate source files.
2055 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
2058 *) Use a less unusual form of the Miller-Rabin primality test (it used
2059 a binary algorithm for exponentiation integrated into the Miller-Rabin
2060 loop, our standard modexp algorithms are faster).
2063 *) Support for the EBCDIC character set completed.
2064 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
2066 *) Source code cleanups: use const where appropriate, eliminate casts,
2067 use void * instead of char * in lhash.
2070 *) Bugfix: ssl3_send_server_key_exchange was not restartable
2071 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
2072 this the server could overwrite ephemeral keys that the client
2076 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
2077 using 50 iterations of the Rabin-Miller test.
2079 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
2080 iterations of the Rabin-Miller test as required by the appendix
2081 to FIPS PUB 186[-1]) instead of DSA_is_prime.
2082 As BN_is_prime_fasttest includes trial division, DSA parameter
2083 generation becomes much faster.
2085 This implies a change for the callback functions in DSA_is_prime
2086 and DSA_generate_parameters: The callback function is called once
2087 for each positive witness in the Rabin-Miller test, not just
2088 occasionally in the inner loop; and the parameters to the
2089 callback function now provide an iteration count for the outer
2090 loop rather than for the current invocation of the inner loop.
2091 DSA_generate_parameters additionally can call the callback
2092 function with an 'iteration count' of -1, meaning that a
2093 candidate has passed the trial division test (when q is generated
2094 from an application-provided seed, trial division is skipped).
2097 *) New function BN_is_prime_fasttest that optionally does trial
2098 division before starting the Rabin-Miller test and has
2099 an additional BN_CTX * argument (whereas BN_is_prime always
2100 has to allocate at least one BN_CTX).
2101 'callback(1, -1, cb_arg)' is called when a number has passed the
2102 trial division stage.
2105 *) Fix for bug in CRL encoding. The validity dates weren't being handled
2109 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
2112 *) New function BN_pseudo_rand().
2115 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
2116 bignum version of BN_from_montgomery() with the working code from
2117 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
2121 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
2122 made it impossible to use the same SSL_SESSION data structure in
2123 SSL2 clients in multiple threads.
2126 *) The return value of RAND_load_file() no longer counts bytes obtained
2127 by stat(). RAND_load_file(..., -1) is new and uses the complete file
2128 to seed the PRNG (previously an explicit byte count was required).
2129 [Ulf Möller, Bodo Möller]
2131 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
2132 used (char *) instead of (void *) and had casts all over the place.
2135 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
2138 *) Retain source code compatibility for BN_prime_checks macro:
2139 BN_is_prime(..., BN_prime_checks, ...) now uses
2140 BN_prime_checks_for_size to determine the appropriate number of
2141 Rabin-Miller iterations.
2144 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
2145 DH_CHECK_P_NOT_SAFE_PRIME.
2146 (Check if this is true? OpenPGP calls them "strong".)
2149 *) Merge the functionality of "dh" and "gendh" programs into a new program
2150 "dhparam". The old programs are retained for now but will handle DH keys
2151 (instead of parameters) in future.
2154 *) Make the ciphers, s_server and s_client programs check the return values
2155 when a new cipher list is set.
2158 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
2159 ciphers. Before when the 56bit ciphers were enabled the sorting was
2162 The syntax for the cipher sorting has been extended to support sorting by
2163 cipher-strength (using the strength_bits hard coded in the tables).
2164 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
2166 Fix a bug in the cipher-command parser: when supplying a cipher command
2167 string with an "undefined" symbol (neither command nor alphanumeric
2168 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
2169 an error is flagged.
2171 Due to the strength-sorting extension, the code of the
2172 ssl_create_cipher_list() function was completely rearranged. I hope that
2173 the readability was also increased :-)
2174 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
2176 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
2177 for the first serial number and places 2 in the serial number file. This
2178 avoids problems when the root CA is created with serial number zero and
2179 the first user certificate has the same issuer name and serial number
2183 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
2184 the new code. Add documentation for this stuff.
2187 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
2188 X509_*() to X509at_*() on the grounds that they don't handle X509
2189 structures and behave in an analagous way to the X509v3 functions:
2190 they shouldn't be called directly but wrapper functions should be used
2193 So we also now have some wrapper functions that call the X509at functions
2194 when passed certificate requests. (TO DO: similar things can be done with
2195 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
2196 things. Some of these need some d2i or i2d and print functionality
2197 because they handle more complex structures.)
2200 *) Add missing #ifndefs that caused missing symbols when building libssl
2201 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
2202 NO_RSA in ssl/s2*.c.
2203 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
2205 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
2206 has a return value which indicates the quality of the random data
2207 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
2208 error queue. New function RAND_pseudo_bytes() generates output that is
2209 guaranteed to be unique but not unpredictable. RAND_add is like
2210 RAND_seed, but takes an extra argument for an entropy estimate
2211 (RAND_seed always assumes full entropy).
2214 *) Do more iterations of Rabin-Miller probable prime test (specifically,
2215 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
2216 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
2217 in crypto/bn/bn_prime.c for the complete table). This guarantees a
2218 false-positive rate of at most 2^-80 for random input.
2221 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
2224 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
2225 in the 0.9.5 release), this returns the chain
2226 from an X509_CTX structure with a dup of the stack and all
2227 the X509 reference counts upped: so the stack will exist
2228 after X509_CTX_cleanup() has been called. Modify pkcs12.c
2231 Also make SSL_SESSION_print() print out the verify return
2235 *) Add manpage for the pkcs12 command. Also change the default
2236 behaviour so MAC iteration counts are used unless the new
2237 -nomaciter option is used. This improves file security and
2238 only older versions of MSIE (4.0 for example) need it.
2241 *) Honor the no-xxx Configure options when creating .DEF files.
2244 *) Add PKCS#10 attributes to field table: challengePassword,
2245 unstructuredName and unstructuredAddress. These are taken from
2246 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
2247 international characters are used.
2249 More changes to X509_ATTRIBUTE code: allow the setting of types
2250 based on strings. Remove the 'loc' parameter when adding
2251 attributes because these will be a SET OF encoding which is sorted
2255 *) Initial changes to the 'req' utility to allow request generation
2256 automation. This will allow an application to just generate a template
2257 file containing all the field values and have req construct the
2260 Initial support for X509_ATTRIBUTE handling. Stacks of these are
2261 used all over the place including certificate requests and PKCS#7
2262 structures. They are currently handled manually where necessary with
2263 some primitive wrappers for PKCS#7. The new functions behave in a
2264 manner analogous to the X509 extension functions: they allow
2265 attributes to be looked up by NID and added.
2267 Later something similar to the X509V3 code would be desirable to
2268 automatically handle the encoding, decoding and printing of the
2269 more complex types. The string types like challengePassword can
2270 be handled by the string table functions.
2272 Also modified the multi byte string table handling. Now there is
2273 a 'global mask' which masks out certain types. The table itself
2274 can use the flag STABLE_NO_MASK to ignore the mask setting: this
2275 is useful when for example there is only one permissible type
2276 (as in countryName) and using the mask might result in no valid
2280 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
2281 SSL_get_peer_finished to allow applications to obtain the latest
2282 Finished messages sent to the peer or expected from the peer,
2283 respectively. (SSL_get_peer_finished is usually the Finished message
2284 actually received from the peer, otherwise the protocol will be aborted.)
2286 As the Finished message are message digests of the complete handshake
2287 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
2288 be used for external authentication procedures when the authentication
2289 provided by SSL/TLS is not desired or is not enough.
2292 *) Enhanced support for Alpha Linux is added. Now ./config checks if
2293 the host supports BWX extension and if Compaq C is present on the
2294 $PATH. Just exploiting of the BWX extension results in 20-30%
2295 performance kick for some algorithms, e.g. DES and RC4 to mention
2296 a couple. Compaq C in turn generates ~20% faster code for MD5 and
2300 *) Add support for MS "fast SGC". This is arguably a violation of the
2301 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
2302 weak crypto and after checking the certificate is SGC a second one
2303 with strong crypto. MS SGC stops the first handshake after receiving
2304 the server certificate message and sends a second client hello. Since
2305 a server will typically do all the time consuming operations before
2306 expecting any further messages from the client (server key exchange
2307 is the most expensive) there is little difference between the two.
2309 To get OpenSSL to support MS SGC we have to permit a second client
2310 hello message after we have sent server done. In addition we have to
2311 reset the MAC if we do get this second client hello.
2314 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
2315 if a DER encoded private key is RSA or DSA traditional format. Changed
2316 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
2317 format DER encoded private key. Newer code should use PKCS#8 format which
2318 has the key type encoded in the ASN1 structure. Added DER private key
2319 support to pkcs8 application.
2322 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
2323 ciphersuites has been selected (as required by the SSL 3/TLS 1
2324 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
2325 is set, we interpret this as a request to violate the specification
2326 (the worst that can happen is a handshake failure, and 'correct'
2327 behaviour would result in a handshake failure anyway).
2330 *) In SSL_CTX_add_session, take into account that there might be multiple
2331 SSL_SESSION structures with the same session ID (e.g. when two threads
2332 concurrently obtain them from an external cache).
2333 The internal cache can handle only one SSL_SESSION with a given ID,
2334 so if there's a conflict, we now throw out the old one to achieve
2338 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
2339 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
2340 some routines that use cipher OIDs: some ciphers do not have OIDs
2341 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
2345 *) Simplify the trust setting structure and code. Now we just have
2346 two sequences of OIDs for trusted and rejected settings. These will
2347 typically have values the same as the extended key usage extension
2348 and any application specific purposes.
2350 The trust checking code now has a default behaviour: it will just
2351 check for an object with the same NID as the passed id. Functions can
2352 be provided to override either the default behaviour or the behaviour
2353 for a given id. SSL client, server and email already have functions
2354 in place for compatibility: they check the NID and also return "trusted"
2355 if the certificate is self signed.
2358 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
2359 traditional format into an EVP_PKEY structure.
2362 *) Add a password callback function PEM_cb() which either prompts for
2363 a password if usr_data is NULL or otherwise assumes it is a null
2364 terminated password. Allow passwords to be passed on command line
2365 environment or config files in a few more utilities.
2368 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
2369 keys. Add some short names for PKCS#8 PBE algorithms and allow them
2370 to be specified on the command line for the pkcs8 and pkcs12 utilities.
2371 Update documentation.
2374 *) Support for ASN1 "NULL" type. This could be handled before by using
2375 ASN1_TYPE but there wasn't any function that would try to read a NULL
2376 and produce an error if it couldn't. For compatibility we also have
2377 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
2378 don't allocate anything because they don't need to.
2381 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
2383 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
2385 *) Rebuild of the memory allocation routines used by OpenSSL code and
2386 possibly others as well. The purpose is to make an interface that
2387 provide hooks so anyone can build a separate set of allocation and
2388 deallocation routines to be used by OpenSSL, for example memory
2389 pool implementations, or something else, which was previously hard
2390 since Malloc(), Realloc() and Free() were defined as macros having
2391 the values malloc, realloc and free, respectively (except for Win32
2392 compilations). The same is provided for memory debugging code.
2393 OpenSSL already comes with functionality to find memory leaks, but
2394 this gives people a chance to debug other memory problems.
2396 With these changes, a new set of functions and macros have appeared:
2398 CRYPTO_set_mem_debug_functions() [F]
2399 CRYPTO_get_mem_debug_functions() [F]
2400 CRYPTO_dbg_set_options() [F]
2401 CRYPTO_dbg_get_options() [F]
2402 CRYPTO_malloc_debug_init() [M]
2404 The memory debug functions are NULL by default, unless the library
2405 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
2406 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
2407 gives the standard debugging functions that come with OpenSSL) or
2408 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
2409 provided by the library user) must be used. When the standard
2410 debugging functions are used, CRYPTO_dbg_set_options can be used to
2411 request additional information:
2412 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
2413 the CRYPTO_MDEBUG_xxx macro when compiling the library.
2415 Also, things like CRYPTO_set_mem_functions will always give the
2416 expected result (the new set of functions is used for allocation
2417 and deallocation) at all times, regardless of platform and compiler
2420 To finish it up, some functions that were never use in any other
2421 way than through macros have a new API and new semantic:
2424 CRYPTO_dbg_realloc()
2427 All macros of value have retained their old syntax.
2428 [Richard Levitte and Bodo Moeller]
2430 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
2431 ordering of SMIMECapabilities wasn't in "strength order" and there
2432 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
2436 *) Some ASN1 types with illegal zero length encoding (INTEGER,
2437 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
2438 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
2440 *) Merge in my S/MIME library for OpenSSL. This provides a simple
2441 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
2442 functionality to handle multipart/signed properly) and a utility
2443 called 'smime' to call all this stuff. This is based on code I
2444 originally wrote for Celo who have kindly allowed it to be
2445 included in OpenSSL.
2448 *) Add variants des_set_key_checked and des_set_key_unchecked of
2449 des_set_key (aka des_key_sched). Global variable des_check_key
2450 decides which of these is called by des_set_key; this way
2451 des_check_key behaves as it always did, but applications and
2452 the library itself, which was buggy for des_check_key == 1,
2453 have a cleaner way to pick the version they need.
2456 *) New function PKCS12_newpass() which changes the password of a
2460 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
2461 dynamic mix. In both cases the ids can be used as an index into the
2462 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
2463 functions so they accept a list of the field values and the
2464 application doesn't need to directly manipulate the X509_TRUST
2468 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
2472 *) Modify the way the V3 extension code looks up extensions. This now
2473 works in a similar way to the object code: we have some "standard"
2474 extensions in a static table which is searched with OBJ_bsearch()
2475 and the application can add dynamic ones if needed. The file
2476 crypto/x509v3/ext_dat.h now has the info: this file needs to be
2477 updated whenever a new extension is added to the core code and kept
2478 in ext_nid order. There is a simple program 'tabtest.c' which checks
2479 this. New extensions are not added too often so this file can readily
2480 be maintained manually.
2482 There are two big advantages in doing things this way. The extensions
2483 can be looked up immediately and no longer need to be "added" using
2484 X509V3_add_standard_extensions(): this function now does nothing.
2485 [Side note: I get *lots* of email saying the extension code doesn't
2486 work because people forget to call this function]
2487 Also no dynamic allocation is done unless new extensions are added:
2488 so if we don't add custom extensions there is no need to call
2489 X509V3_EXT_cleanup().
2492 *) Modify enc utility's salting as follows: make salting the default. Add a
2493 magic header, so unsalted files fail gracefully instead of just decrypting
2494 to garbage. This is because not salting is a big security hole, so people
2495 should be discouraged from doing it.
2498 *) Fixes and enhancements to the 'x509' utility. It allowed a message
2499 digest to be passed on the command line but it only used this
2500 parameter when signing a certificate. Modified so all relevant
2501 operations are affected by the digest parameter including the
2502 -fingerprint and -x509toreq options. Also -x509toreq choked if a
2503 DSA key was used because it didn't fix the digest.
2506 *) Initial certificate chain verify code. Currently tests the untrusted
2507 certificates for consistency with the verify purpose (which is set
2508 when the X509_STORE_CTX structure is set up) and checks the pathlength.
2510 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
2511 this is because it will reject chains with invalid extensions whereas
2512 every previous version of OpenSSL and SSLeay made no checks at all.
2514 Trust code: checks the root CA for the relevant trust settings. Trust
2515 settings have an initial value consistent with the verify purpose: e.g.
2516 if the verify purpose is for SSL client use it expects the CA to be
2517 trusted for SSL client use. However the default value can be changed to
2518 permit custom trust settings: one example of this would be to only trust
2519 certificates from a specific "secure" set of CAs.
2521 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
2522 which should be used for version portability: especially since the
2523 verify structure is likely to change more often now.
2525 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
2526 to set them. If not set then assume SSL clients will verify SSL servers
2529 Two new options to the verify program: -untrusted allows a set of
2530 untrusted certificates to be passed in and -purpose which sets the
2531 intended purpose of the certificate. If a purpose is set then the
2532 new chain verify code is used to check extension consistency.
2535 *) Support for the authority information access extension.
2538 *) Modify RSA and DSA PEM read routines to transparently handle
2539 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
2540 public keys in a format compatible with certificate
2541 SubjectPublicKeyInfo structures. Unfortunately there were already
2542 functions called *_PublicKey_* which used various odd formats so
2543 these are retained for compatibility: however the DSA variants were
2544 never in a public release so they have been deleted. Changed dsa/rsa
2545 utilities to handle the new format: note no releases ever handled public
2546 keys so we should be OK.
2548 The primary motivation for this change is to avoid the same fiasco
2549 that dogs private keys: there are several incompatible private key
2550 formats some of which are standard and some OpenSSL specific and
2551 require various evil hacks to allow partial transparent handling and
2552 even then it doesn't work with DER formats. Given the option anything
2553 other than PKCS#8 should be dumped: but the other formats have to
2554 stay in the name of compatibility.
2556 With public keys and the benefit of hindsight one standard format
2557 is used which works with EVP_PKEY, RSA or DSA structures: though
2558 it clearly returns an error if you try to read the wrong kind of key.
2560 Added a -pubkey option to the 'x509' utility to output the public key.
2561 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
2562 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
2563 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
2564 that do the same as the EVP_PKEY_assign_*() except they up the
2565 reference count of the added key (they don't "swallow" the
2569 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
2570 CRLs would fail if the file contained no certificates or no CRLs:
2571 added a new function to read in both types and return the number
2572 read: this means that if none are read it will be an error. The
2573 DER versions of the certificate and CRL reader would always fail
2574 because it isn't possible to mix certificates and CRLs in DER format
2575 without choking one or the other routine. Changed this to just read
2576 a certificate: this is the best we can do. Also modified the code
2577 in apps/verify.c to take notice of return codes: it was previously
2578 attempting to read in certificates from NULL pointers and ignoring
2579 any errors: this is one reason why the cert and CRL reader seemed
2580 to work. It doesn't check return codes from the default certificate
2581 routines: these may well fail if the certificates aren't installed.
2584 *) Code to support otherName option in GeneralName.
2587 *) First update to verify code. Change the verify utility
2588 so it warns if it is passed a self signed certificate:
2589 for consistency with the normal behaviour. X509_verify
2590 has been modified to it will now verify a self signed
2591 certificate if *exactly* the same certificate appears
2592 in the store: it was previously impossible to trust a
2593 single self signed certificate. This means that:
2594 openssl verify ss.pem
2595 now gives a warning about a self signed certificate but
2596 openssl verify -CAfile ss.pem ss.pem
2600 *) For servers, store verify_result in SSL_SESSION data structure
2601 (and add it to external session representation).
2602 This is needed when client certificate verifications fails,
2603 but an application-provided verification callback (set by
2604 SSL_CTX_set_cert_verify_callback) allows accepting the session
2605 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
2606 but returns 1): When the session is reused, we have to set
2607 ssl->verify_result to the appropriate error code to avoid
2609 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
2611 *) Fix a bug in the new PKCS#7 code: it didn't consider the
2612 case in PKCS7_dataInit() where the signed PKCS7 structure
2613 didn't contain any existing data because it was being created.
2614 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
2616 *) Add a salt to the key derivation routines in enc.c. This
2617 forms the first 8 bytes of the encrypted file. Also add a
2618 -S option to allow a salt to be input on the command line.
2621 *) New function X509_cmp(). Oddly enough there wasn't a function
2622 to compare two certificates. We do this by working out the SHA1
2623 hash and comparing that. X509_cmp() will be needed by the trust
2627 *) SSL_get1_session() is like SSL_get_session(), but increments
2628 the reference count in the SSL_SESSION returned.
2629 [Geoff Thorpe <geoff@eu.c2.net>]
2631 *) Fix for 'req': it was adding a null to request attributes.
2632 Also change the X509_LOOKUP and X509_INFO code to handle
2633 certificate auxiliary information.
2636 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
2640 *) Add the possibility to add extra information to the memory leak
2641 detecting output, to form tracebacks, showing from where each
2642 allocation was originated: CRYPTO_push_info("constant string") adds
2643 the string plus current file name and line number to a per-thread
2644 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
2645 is like calling CYRPTO_pop_info() until the stack is empty.
2646 Also updated memory leak detection code to be multi-thread-safe.
2649 *) Add options -text and -noout to pkcs7 utility and delete the
2650 encryption options which never did anything. Update docs.
2653 *) Add options to some of the utilities to allow the pass phrase
2654 to be included on either the command line (not recommended on
2655 OSes like Unix) or read from the environment. Update the
2656 manpages and fix a few bugs.
2659 *) Add a few manpages for some of the openssl commands.
2662 *) Fix the -revoke option in ca. It was freeing up memory twice,
2663 leaking and not finding already revoked certificates.
2666 *) Extensive changes to support certificate auxiliary information.
2667 This involves the use of X509_CERT_AUX structure and X509_AUX
2668 functions. An X509_AUX function such as PEM_read_X509_AUX()
2669 can still read in a certificate file in the usual way but it
2670 will also read in any additional "auxiliary information". By
2671 doing things this way a fair degree of compatibility can be
2672 retained: existing certificates can have this information added
2673 using the new 'x509' options.
2675 Current auxiliary information includes an "alias" and some trust
2676 settings. The trust settings will ultimately be used in enhanced
2677 certificate chain verification routines: currently a certificate
2678 can only be trusted if it is self signed and then it is trusted
2682 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
2683 The problem was that one of the replacement routines had not been working
2684 since SSLeay releases. For now the offending routine has been replaced
2685 with non-optimised assembler. Even so, this now gives around 95%
2686 performance improvement for 1024 bit RSA signs.
2689 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
2690 handling. Most clients have the effective key size in bits equal to
2691 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
2692 A few however don't do this and instead use the size of the decrypted key
2693 to determine the RC2 key length and the AlgorithmIdentifier to determine
2694 the effective key length. In this case the effective key length can still
2695 be 40 bits but the key length can be 168 bits for example. This is fixed
2696 by manually forcing an RC2 key into the EVP_PKEY structure because the
2697 EVP code can't currently handle unusual RC2 key sizes: it always assumes
2698 the key length and effective key length are equal.
2701 *) Add a bunch of functions that should simplify the creation of
2702 X509_NAME structures. Now you should be able to do:
2703 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
2704 and have it automatically work out the correct field type and fill in
2705 the structures. The more adventurous can try:
2706 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
2707 and it will (hopefully) work out the correct multibyte encoding.
2710 *) Change the 'req' utility to use the new field handling and multibyte
2711 copy routines. Before the DN field creation was handled in an ad hoc
2712 way in req, ca, and x509 which was rather broken and didn't support
2713 BMPStrings or UTF8Strings. Since some software doesn't implement
2714 BMPStrings or UTF8Strings yet, they can be enabled using the config file
2715 using the dirstring_type option. See the new comment in the default
2716 openssl.cnf for more info.
2719 *) Make crypto/rand/md_rand.c more robust:
2720 - Assure unique random numbers after fork().
2721 - Make sure that concurrent threads access the global counter and
2722 md serializably so that we never lose entropy in them
2723 or use exactly the same state in multiple threads.
2724 Access to the large state is not always serializable because
2725 the additional locking could be a performance killer, and
2726 md should be large enough anyway.
2729 *) New file apps/app_rand.c with commonly needed functionality
2730 for handling the random seed file.
2732 Use the random seed file in some applications that previously did not:
2734 dsaparam -genkey (which also ignored its '-rand' option),
2737 x509 (when signing).
2738 Except on systems with /dev/urandom, it is crucial to have a random
2739 seed file at least for key creation, DSA signing, and for DH exchanges;
2740 for RSA signatures we could do without one.
2742 gendh and gendsa (unlike genrsa) used to read only the first byte
2743 of each file listed in the '-rand' option. The function as previously
2744 found in genrsa is now in app_rand.c and is used by all programs
2745 that support '-rand'.
2748 *) In RAND_write_file, use mode 0600 for creating files;
2749 don't just chmod when it may be too late.
2752 *) Report an error from X509_STORE_load_locations
2753 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
2756 *) New function ASN1_mbstring_copy() this copies a string in either
2757 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
2758 into an ASN1_STRING type. A mask of permissible types is passed
2759 and it chooses the "minimal" type to use or an error if not type
2763 *) Add function equivalents to the various macros in asn1.h. The old
2764 macros are retained with an M_ prefix. Code inside the library can
2765 use the M_ macros. External code (including the openssl utility)
2766 should *NOT* in order to be "shared library friendly".
2769 *) Add various functions that can check a certificate's extensions
2770 to see if it usable for various purposes such as SSL client,
2771 server or S/MIME and CAs of these types. This is currently
2772 VERY EXPERIMENTAL but will ultimately be used for certificate chain
2773 verification. Also added a -purpose flag to x509 utility to
2774 print out all the purposes.
2777 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
2781 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
2782 for, obtain and decode and extension and obtain its critical flag.
2783 This allows all the necessary extension code to be handled in a
2784 single function call.
2787 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
2788 platforms. See crypto/rc4/rc4_enc.c for further details.
2791 *) New -noout option to asn1parse. This causes no output to be produced
2792 its main use is when combined with -strparse and -out to extract data
2793 from a file (which may not be in ASN.1 format).
2796 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
2797 when producing the local key id.
2798 [Richard Levitte <levitte@stacken.kth.se>]
2800 *) New option -dhparam in s_server. This allows a DH parameter file to be
2801 stated explicitly. If it is not stated then it tries the first server
2802 certificate file. The previous behaviour hard coded the filename
2806 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
2807 a public key to be input or output. For example:
2808 openssl rsa -in key.pem -pubout -out pubkey.pem
2809 Also added necessary DSA public key functions to handle this.
2812 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
2813 in the message. This was handled by allowing
2814 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
2815 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
2817 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
2818 to the end of the strings whereas this didn't. This would cause problems
2819 if strings read with d2i_ASN1_bytes() were later modified.
2820 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
2822 *) Fix for base64 decode bug. When a base64 bio reads only one line of
2823 data and it contains EOF it will end up returning an error. This is
2824 caused by input 46 bytes long. The cause is due to the way base64
2825 BIOs find the start of base64 encoded data. They do this by trying a
2826 trial decode on each line until they find one that works. When they
2827 do a flag is set and it starts again knowing it can pass all the
2828 data directly through the decoder. Unfortunately it doesn't reset
2829 the context it uses. This means that if EOF is reached an attempt
2830 is made to pass two EOFs through the context and this causes the
2831 resulting error. This can also cause other problems as well. As is
2832 usual with these problems it takes *ages* to find and the fix is
2833 trivial: move one line.
2834 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
2836 *) Ugly workaround to get s_client and s_server working under Windows. The
2837 old code wouldn't work because it needed to select() on sockets and the
2838 tty (for keypresses and to see if data could be written). Win32 only
2839 supports select() on sockets so we select() with a 1s timeout on the
2840 sockets and then see if any characters are waiting to be read, if none
2841 are present then we retry, we also assume we can always write data to
2842 the tty. This isn't nice because the code then blocks until we've
2843 received a complete line of data and it is effectively polling the
2844 keyboard at 1s intervals: however it's quite a bit better than not
2845 working at all :-) A dedicated Windows application might handle this
2846 with an event loop for example.
2849 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
2850 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
2851 will be called when RSA_sign() and RSA_verify() are used. This is useful
2852 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
2853 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
2854 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
2855 This necessitated the support of an extra signature type NID_md5_sha1
2856 for SSL signatures and modifications to the SSL library to use it instead
2857 of calling RSA_public_decrypt() and RSA_private_encrypt().
2860 *) Add new -verify -CAfile and -CApath options to the crl program, these
2861 will lookup a CRL issuers certificate and verify the signature in a
2862 similar way to the verify program. Tidy up the crl program so it
2863 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
2864 less strict. It will now permit CRL extensions even if it is not
2865 a V2 CRL: this will allow it to tolerate some broken CRLs.
2868 *) Initialize all non-automatic variables each time one of the openssl
2869 sub-programs is started (this is necessary as they may be started
2870 multiple times from the "OpenSSL>" prompt).
2871 [Lennart Bang, Bodo Moeller]
2873 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
2874 removing all other RSA functionality (this is what NO_RSA does). This
2875 is so (for example) those in the US can disable those operations covered
2876 by the RSA patent while allowing storage and parsing of RSA keys and RSA
2880 *) Non-copying interface to BIO pairs.
2881 (still largely untested)
2884 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
2885 ASCII string. This was handled independently in various places before.
2888 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
2889 UTF8 strings a character at a time.
2892 *) Use client_version from client hello to select the protocol
2893 (s23_srvr.c) and for RSA client key exchange verification
2894 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
2897 *) Add various utility functions to handle SPKACs, these were previously
2898 handled by poking round in the structure internals. Added new function
2899 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
2900 print, verify and generate SPKACs. Based on an original idea from
2901 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
2904 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
2907 *) Allow the config file extension section to be overwritten on the
2908 command line. Based on an original idea from Massimiliano Pala
2909 <madwolf@comune.modena.it>. The new option is called -extensions
2910 and can be applied to ca, req and x509. Also -reqexts to override
2911 the request extensions in req and -crlexts to override the crl extensions
2915 *) Add new feature to the SPKAC handling in ca. Now you can include
2916 the same field multiple times by preceding it by "XXXX." for example:
2919 this is the same syntax as used in the req config file.
2922 *) Allow certificate extensions to be added to certificate requests. These
2923 are specified in a 'req_extensions' option of the req section of the
2924 config file. They can be printed out with the -text option to req but
2925 are otherwise ignored at present.
2928 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
2929 data read consists of only the final block it would not decrypted because
2930 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
2931 A misplaced 'break' also meant the decrypted final block might not be
2932 copied until the next read.
2935 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
2936 a few extra parameters to the DH structure: these will be useful if
2937 for example we want the value of 'q' or implement X9.42 DH.
2940 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
2941 provides hooks that allow the default DSA functions or functions on a
2942 "per key" basis to be replaced. This allows hardware acceleration and
2943 hardware key storage to be handled without major modification to the
2944 library. Also added low level modexp hooks and CRYPTO_EX structure and
2945 associated functions.
2948 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
2949 as "read only": it can't be written to and the buffer it points to will
2950 not be freed. Reading from a read only BIO is much more efficient than
2951 a normal memory BIO. This was added because there are several times when
2952 an area of memory needs to be read from a BIO. The previous method was
2953 to create a memory BIO and write the data to it, this results in two
2954 copies of the data and an O(n^2) reading algorithm. There is a new
2955 function BIO_new_mem_buf() which creates a read only memory BIO from
2956 an area of memory. Also modified the PKCS#7 routines to use read only
2960 *) Bugfix: ssl23_get_client_hello did not work properly when called in
2961 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
2962 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
2963 but a retry condition occured while trying to read the rest.
2966 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
2967 NID_pkcs7_encrypted by default: this was wrong since this should almost
2968 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
2969 the encrypted data type: this is a more sensible place to put it and it
2970 allows the PKCS#12 code to be tidied up that duplicated this
2974 *) Changed obj_dat.pl script so it takes its input and output files on
2975 the command line. This should avoid shell escape redirection problems
2979 *) Initial support for certificate extension requests, these are included
2980 in things like Xenroll certificate requests. Included functions to allow
2981 extensions to be obtained and added.
2984 *) -crlf option to s_client and s_server for sending newlines as
2985 CRLF (as required by many protocols).
2988 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
2990 *) Install libRSAglue.a when OpenSSL is built with RSAref.
2991 [Ralf S. Engelschall]
2993 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
2994 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
2996 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
3000 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
3001 DH parameters/keys (q is lost during that conversion, but the resulting
3002 DH parameters contain its length).
3004 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
3005 much faster than DH_generate_parameters (which creates parameters
3006 where p = 2*q + 1), and also the smaller q makes DH computations
3007 much more efficient (160-bit exponentiation instead of 1024-bit
3008 exponentiation); so this provides a convenient way to support DHE
3009 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
3010 utter importance to use
3011 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
3013 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
3014 when such DH parameters are used, because otherwise small subgroup
3015 attacks may become possible!
3018 *) Avoid memory leak in i2d_DHparams.
3021 *) Allow the -k option to be used more than once in the enc program:
3022 this allows the same encrypted message to be read by multiple recipients.
3025 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
3026 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
3027 it will always use the numerical form of the OID, even if it has a short
3031 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
3032 method only got called if p,q,dmp1,dmq1,iqmp components were present,
3033 otherwise bn_mod_exp was called. In the case of hardware keys for example
3034 no private key components need be present and it might store extra data
3035 in the RSA structure, which cannot be accessed from bn_mod_exp.
3036 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
3037 private key operations.
3040 *) Added support for SPARC Linux.
3043 *) pem_password_cb function type incompatibly changed from
3044 typedef int pem_password_cb(char *buf, int size, int rwflag);
3046 ....(char *buf, int size, int rwflag, void *userdata);
3047 so that applications can pass data to their callbacks:
3048 The PEM[_ASN1]_{read,write}... functions and macros now take an
3049 additional void * argument, which is just handed through whenever
3050 the password callback is called.
3051 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
3053 New function SSL_CTX_set_default_passwd_cb_userdata.
3055 Compatibility note: As many C implementations push function arguments
3056 onto the stack in reverse order, the new library version is likely to
3057 interoperate with programs that have been compiled with the old
3058 pem_password_cb definition (PEM_whatever takes some data that
3059 happens to be on the stack as its last argument, and the callback
3060 just ignores this garbage); but there is no guarantee whatsoever that
3063 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
3064 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
3065 problems not only on Windows, but also on some Unix platforms.
3066 To avoid problematic command lines, these definitions are now in an
3067 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
3068 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
3071 *) MIPS III/IV assembler module is reimplemented.
3074 *) More DES library cleanups: remove references to srand/rand and
3075 delete an unused file.
3078 *) Add support for the the free Netwide assembler (NASM) under Win32,
3079 since not many people have MASM (ml) and it can be hard to obtain.
3080 This is currently experimental but it seems to work OK and pass all
3081 the tests. Check out INSTALL.W32 for info.
3084 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
3085 without temporary keys kept an extra copy of the server key,
3086 and connections with temporary keys did not free everything in case
3090 *) New function RSA_check_key and new openssl rsa option -check
3091 for verifying the consistency of RSA keys.
3092 [Ulf Moeller, Bodo Moeller]
3094 *) Various changes to make Win32 compile work:
3095 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
3096 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
3097 comparison" warnings.
3098 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
3101 *) Add a debugging option to PKCS#5 v2 key generation function: when
3102 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
3103 derived keys are printed to stderr.
3106 *) Copy the flags in ASN1_STRING_dup().
3107 [Roman E. Pavlov <pre@mo.msk.ru>]
3109 *) The x509 application mishandled signing requests containing DSA
3110 keys when the signing key was also DSA and the parameters didn't match.
3112 It was supposed to omit the parameters when they matched the signing key:
3113 the verifying software was then supposed to automatically use the CA's
3114 parameters if they were absent from the end user certificate.
3116 Omitting parameters is no longer recommended. The test was also
3117 the wrong way round! This was probably due to unusual behaviour in
3118 EVP_cmp_parameters() which returns 1 if the parameters match.
3119 This meant that parameters were omitted when they *didn't* match and
3120 the certificate was useless. Certificates signed with 'ca' didn't have
3122 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
3124 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
3125 The interface is as follows:
3126 Applications can use
3127 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
3128 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
3129 "off" is now the default.
3130 The library internally uses
3131 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
3132 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
3133 to disable memory-checking temporarily.
3135 Some inconsistent states that previously were possible (and were
3136 even the default) are now avoided.
3138 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
3139 with each memory chunk allocated; this is occasionally more helpful
3140 than just having a counter.
3142 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
3144 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
3148 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
3149 which largely parallels "options", but is for changing API behaviour,
3150 whereas "options" are about protocol behaviour.
3151 Initial "mode" flags are:
3153 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
3154 a single record has been written.
3155 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
3156 retries use the same buffer location.
3157 (But all of the contents must be
3161 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
3164 *) Fix problems with no-hmac etc.
3165 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
3167 *) New functions RSA_get_default_method(), RSA_set_method() and
3168 RSA_get_method(). These allows replacement of RSA_METHODs without having
3169 to mess around with the internals of an RSA structure.
3172 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
3173 Also really enable memory leak checks in openssl.c and in some
3175 [Chad C. Mulligan, Bodo Moeller]
3177 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
3178 up the length of negative integers. This has now been simplified to just
3179 store the length when it is first determined and use it later, rather
3180 than trying to keep track of where data is copied and updating it to
3182 [Steve Henson, reported by Brien Wheeler
3183 <bwheeler@authentica-security.com>]
3185 *) Add a new function PKCS7_signatureVerify. This allows the verification
3186 of a PKCS#7 signature but with the signing certificate passed to the
3187 function itself. This contrasts with PKCS7_dataVerify which assumes the
3188 certificate is present in the PKCS#7 structure. This isn't always the
3189 case: certificates can be omitted from a PKCS#7 structure and be
3190 distributed by "out of band" means (such as a certificate database).
3193 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
3194 function prototypes in pem.h, also change util/mkdef.pl to add the
3195 necessary function names.
3198 *) mk1mf.pl (used by Windows builds) did not properly read the
3199 options set by Configure in the top level Makefile, and Configure
3200 was not even able to write more than one option correctly.
3201 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
3204 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
3205 file to be loaded from a BIO or FILE pointer. The BIO version will
3206 for example allow memory BIOs to contain config info.
3209 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
3210 Whoever hopes to achieve shared-library compatibility across versions
3211 must use this, not the compile-time macro.
3212 (Exercise 0.9.4: Which is the minimum library version required by
3214 Note: All this applies only to multi-threaded programs, others don't
3218 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
3219 through a BIO pair triggered the default case, i.e.
3220 SSLerr(...,SSL_R_UNKNOWN_STATE).
3223 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
3224 can use the SSL library even if none of the specific BIOs is
3228 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
3229 for the encoded length.
3230 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
3232 *) Add initial documentation of the X509V3 functions.
3235 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
3236 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
3237 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
3238 secure PKCS#8 private key format with a high iteration count.
3241 *) Fix determination of Perl interpreter: A perl or perl5
3242 _directory_ in $PATH was also accepted as the interpreter.
3243 [Ralf S. Engelschall]
3245 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
3246 wrong with it but it was very old and did things like calling
3247 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
3251 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
3252 to use the new extension code.
3255 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
3256 with macros. This should make it easier to change their form, add extra
3257 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
3261 *) Add to configuration table a new entry that can specify an alternative
3262 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
3263 according to Mark Crispin <MRC@Panda.COM>.
3267 *) DES CBC did not update the IV. Weird.
3270 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
3271 Changing the behaviour of the former might break existing programs --
3272 where IV updating is needed, des_ncbc_encrypt can be used.
3275 *) When bntest is run from "make test" it drives bc to check its
3276 calculations, as well as internally checking them. If an internal check
3277 fails, it needs to cause bc to give a non-zero result or make test carries
3278 on without noticing the failure. Fixed.
3281 *) DES library cleanups.
3284 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
3285 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
3286 ciphers. NOTE: although the key derivation function has been verified
3287 against some published test vectors it has not been extensively tested
3288 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
3292 *) Instead of "mkdir -p", which is not fully portable, use new
3293 Perl script "util/mkdir-p.pl".
3296 *) Rewrite the way password based encryption (PBE) is handled. It used to
3297 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
3298 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
3299 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
3300 the 'parameter' field of the AlgorithmIdentifier is passed to the
3301 underlying key generation function so it must do its own ASN1 parsing.
3302 This has also changed the EVP_PBE_CipherInit() function which now has a
3303 'parameter' argument instead of literal salt and iteration count values
3304 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
3307 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
3308 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
3309 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
3310 KEY" because this clashed with PKCS#8 unencrypted string. Since this
3311 value was just used as a "magic string" and not used directly its
3312 value doesn't matter.
3315 *) Introduce some semblance of const correctness to BN. Shame C doesn't
3319 *) "linux-sparc64" configuration (ultrapenguin).
3320 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
3321 "linux-sparc" configuration.
3322 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
3324 *) config now generates no-xxx options for missing ciphers.
3327 *) Support the EBCDIC character set (work in progress).
3328 File ebcdic.c not yet included because it has a different license.
3329 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
3331 *) Support BS2000/OSD-POSIX.
3332 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
3334 *) Make callbacks for key generation use void * instead of char *.
3337 *) Make S/MIME samples compile (not yet tested).
3340 *) Additional typesafe stacks.
3343 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
3347 Changes between 0.9.3 and 0.9.3a [29 May 1999]
3349 *) New configuration variant "sco5-gcc".
3351 *) Updated some demos.
3352 [Sean O Riordain, Wade Scholine]
3354 *) Add missing BIO_free at exit of pkcs12 application.
3357 *) Fix memory leak in conf.c.
3360 *) Updates for Win32 to assembler version of MD5.
3363 *) Set #! path to perl in apps/der_chop to where we found it
3364 instead of using a fixed path.
3367 *) SHA library changes for irix64-mips4-cc.
3370 *) Improvements for VMS support.
3374 Changes between 0.9.2b and 0.9.3 [24 May 1999]
3376 *) Bignum library bug fix. IRIX 6 passes "make test" now!
3377 This also avoids the problems with SC4.2 and unpatched SC5.
3378 [Andy Polyakov <appro@fy.chalmers.se>]
3380 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
3381 These are required because of the typesafe stack would otherwise break
3382 existing code. If old code used a structure member which used to be STACK
3383 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
3384 sk_num or sk_value it would produce an error because the num, data members
3385 are not present in STACK_OF. Now it just produces a warning. sk_set
3386 replaces the old method of assigning a value to sk_value
3387 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
3388 that does this will no longer work (and should use sk_set instead) but
3389 this could be regarded as a "questionable" behaviour anyway.
3392 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
3393 correctly handle encrypted S/MIME data.
3396 *) Change type of various DES function arguments from des_cblock
3397 (which means, in function argument declarations, pointer to char)
3398 to des_cblock * (meaning pointer to array with 8 char elements),
3399 which allows the compiler to do more typechecking; it was like
3400 that back in SSLeay, but with lots of ugly casts.
3402 Introduce new type const_des_cblock.
3405 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
3406 problems: find RecipientInfo structure that matches recipient certificate
3407 and initialise the ASN1 structures properly based on passed cipher.
3410 *) Belatedly make the BN tests actually check the results.
3413 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
3414 to and from BNs: it was completely broken. New compilation option
3415 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
3416 key elements as negative integers.
3419 *) Reorganize and speed up MD5.
3420 [Andy Polyakov <appro@fy.chalmers.se>]
3423 [Richard Levitte <richard@levitte.org>]
3425 *) New option -out to asn1parse to allow the parsed structure to be
3426 output to a file. This is most useful when combined with the -strparse
3427 option to examine the output of things like OCTET STRINGS.
3430 *) Make SSL library a little more fool-proof by not requiring any longer
3431 that SSL_set_{accept,connect}_state be called before
3432 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
3433 in many applications because usually everything *appeared* to work as
3434 intended anyway -- now it really works as intended).
3437 *) Move openssl.cnf out of lib/.
3440 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
3441 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
3442 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
3443 [Ralf S. Engelschall]
3445 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
3446 handle PKCS#7 enveloped data properly.
3447 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
3449 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
3450 copying pointers. The cert_st handling is changed by this in
3451 various ways (and thus what used to be known as ctx->default_cert
3452 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
3453 any longer when s->cert does not give us what we need).
3454 ssl_cert_instantiate becomes obsolete by this change.
3455 As soon as we've got the new code right (possibly it already is?),
3456 we have solved a couple of bugs of the earlier code where s->cert
3457 was used as if it could not have been shared with other SSL structures.
3459 Note that using the SSL API in certain dirty ways now will result
3460 in different behaviour than observed with earlier library versions:
3461 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
3462 does not influence s as it used to.
3464 In order to clean up things more thoroughly, inside SSL_SESSION
3465 we don't use CERT any longer, but a new structure SESS_CERT
3466 that holds per-session data (if available); currently, this is
3467 the peer's certificate chain and, for clients, the server's certificate
3468 and temporary key. CERT holds only those values that can have
3469 meaningful defaults in an SSL_CTX.
3472 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
3473 from the internal representation. Various PKCS#7 fixes: remove some
3474 evil casts and set the enc_dig_alg field properly based on the signing
3478 *) Allow PKCS#12 password to be set from the command line or the
3479 environment. Let 'ca' get its config file name from the environment
3480 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
3484 *) Allow certificate policies extension to use an IA5STRING for the
3485 organization field. This is contrary to the PKIX definition but
3486 VeriSign uses it and IE5 only recognises this form. Document 'x509'
3490 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
3491 without disallowing inline assembler and the like for non-pedantic builds.
3494 *) Support Borland C++ builder.
3495 [Janez Jere <jj@void.si>, modified by Ulf Möller]
3500 *) SHA-1 cleanups and performance enhancements.
3501 [Andy Polyakov <appro@fy.chalmers.se>]
3503 *) Sparc v8plus assembler for the bignum library.
3504 [Andy Polyakov <appro@fy.chalmers.se>]
3506 *) Accept any -xxx and +xxx compiler options in Configure.
3509 *) Update HPUX configuration.
3512 *) Add missing sk_<type>_unshift() function to safestack.h
3513 [Ralf S. Engelschall]
3515 *) New function SSL_CTX_use_certificate_chain_file that sets the
3516 "extra_cert"s in addition to the certificate. (This makes sense
3517 only for "PEM" format files, as chains as a whole are not
3521 *) Support verify_depth from the SSL API.
3522 x509_vfy.c had what can be considered an off-by-one-error:
3523 Its depth (which was not part of the external interface)
3524 was actually counting the number of certificates in a chain;
3525 now it really counts the depth.
3528 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
3529 instead of X509err, which often resulted in confusing error
3530 messages since the error codes are not globally unique
3531 (e.g. an alleged error in ssl3_accept when a certificate
3532 didn't match the private key).
3534 *) New function SSL_CTX_set_session_id_context that allows to set a default
3535 value (so that you don't need SSL_set_session_id_context for each
3536 connection using the SSL_CTX).
3539 *) OAEP decoding bug fix.
3542 *) Support INSTALL_PREFIX for package builders, as proposed by
3546 *) New Configure options "threads" and "no-threads". For systems
3547 where the proper compiler options are known (currently Solaris
3548 and Linux), "threads" is the default.
3551 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
3554 *) Install various scripts to $(OPENSSLDIR)/misc, not to
3555 $(INSTALLTOP)/bin -- they shouldn't clutter directories
3556 such as /usr/local/bin.
3559 *) "make linux-shared" to build shared libraries.
3560 [Niels Poppe <niels@netbox.org>]
3562 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
3565 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
3566 extension adding in x509 utility.
3569 *) Remove NOPROTO sections and error code comments.
3572 *) Partial rewrite of the DEF file generator to now parse the ANSI
3576 *) New Configure options --prefix=DIR and --openssldir=DIR.
3579 *) Complete rewrite of the error code script(s). It is all now handled
3580 by one script at the top level which handles error code gathering,
3581 header rewriting and C source file generation. It should be much better
3582 than the old method: it now uses a modified version of Ulf's parser to
3583 read the ANSI prototypes in all header files (thus the old K&R definitions
3584 aren't needed for error creation any more) and do a better job of
3585 translating function codes into names. The old 'ASN1 error code imbedded
3586 in a comment' is no longer necessary and it doesn't use .err files which
3587 have now been deleted. Also the error code call doesn't have to appear all
3588 on one line (which resulted in some large lines...).
3591 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
3594 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
3595 0 (which usually indicates a closed connection), but continue reading.
3598 *) Fix some race conditions.
3601 *) Add support for CRL distribution points extension. Add Certificate
3602 Policies and CRL distribution points documentation.
3605 *) Move the autogenerated header file parts to crypto/opensslconf.h.
3608 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
3609 8 of keying material. Merlin has also confirmed interop with this fix
3610 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
3611 [Merlin Hughes <merlin@baltimore.ie>]
3613 *) Fix lots of warnings.
3614 [Richard Levitte <levitte@stacken.kth.se>]
3616 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
3617 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
3618 [Richard Levitte <levitte@stacken.kth.se>]
3620 *) Fix problems with sizeof(long) == 8.
3621 [Andy Polyakov <appro@fy.chalmers.se>]
3623 *) Change functions to ANSI C.
3626 *) Fix typos in error codes.
3627 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
3629 *) Remove defunct assembler files from Configure.
3632 *) SPARC v8 assembler BIGNUM implementation.
3633 [Andy Polyakov <appro@fy.chalmers.se>]
3635 *) Support for Certificate Policies extension: both print and set.
3636 Various additions to support the r2i method this uses.
3639 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
3640 return a const string when you are expecting an allocated buffer.
3643 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
3644 types DirectoryString and DisplayText.
3647 *) Add code to allow r2i extensions to access the configuration database,
3648 add an LHASH database driver and add several ctx helper functions.
3651 *) Fix an evil bug in bn_expand2() which caused various BN functions to
3652 fail when they extended the size of a BIGNUM.
3655 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
3656 support typesafe stack.
3659 *) Fix typo in SSL_[gs]et_options().
3660 [Nils Frostberg <nils@medcom.se>]
3662 *) Delete various functions and files that belonged to the (now obsolete)
3663 old X509V3 handling code.
3666 *) New Configure option "rsaref".
3669 *) Don't auto-generate pem.h.
3672 *) Introduce type-safe ASN.1 SETs.
3675 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
3676 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
3678 *) Introduce type-safe STACKs. This will almost certainly break lots of code
3679 that links with OpenSSL (well at least cause lots of warnings), but fear
3680 not: the conversion is trivial, and it eliminates loads of evil casts. A
3681 few STACKed things have been converted already. Feel free to convert more.
3682 In the fullness of time, I'll do away with the STACK type altogether.
3685 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
3686 specified in <certfile> by updating the entry in the index.txt file.
3687 This way one no longer has to edit the index.txt file manually for
3688 revoking a certificate. The -revoke option does the gory details now.
3689 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
3691 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
3692 `-text' option at all and this way the `-noout -text' combination was
3693 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
3694 [Ralf S. Engelschall]
3696 *) Make sure a corresponding plain text error message exists for the
3697 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
3698 verify callback function determined that a certificate was revoked.
3699 [Ralf S. Engelschall]
3701 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
3702 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
3703 all available cipers including rc5, which was forgotten until now.
3704 In order to let the testing shell script know which algorithms
3705 are available, a new (up to now undocumented) command
3706 "openssl list-cipher-commands" is used.
3709 *) Bugfix: s_client occasionally would sleep in select() when
3710 it should have checked SSL_pending() first.
3713 *) New functions DSA_do_sign and DSA_do_verify to provide access to
3714 the raw DSA values prior to ASN.1 encoding.
3717 *) Tweaks to Configure
3718 [Niels Poppe <niels@netbox.org>]
3720 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
3724 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
3727 *) New config option to avoid instructions that are illegal on the 80386.
3728 The default code is faster, but requires at least a 486.
3731 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
3732 SSL2_SERVER_VERSION (not used at all) macros, which are now the
3733 same as SSL2_VERSION anyway.
3736 *) New "-showcerts" option for s_client.
3739 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
3740 application. Various cleanups and fixes.
3743 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
3744 modify error routines to work internally. Add error codes and PBE init
3745 to library startup routines.
3748 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
3749 packing functions to asn1 and evp. Changed function names and error
3750 codes along the way.
3753 *) PKCS12 integration: and so it begins... First of several patches to
3754 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
3755 objects to objects.h
3758 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
3759 and display support for Thawte strong extranet extension.
3762 *) Add LinuxPPC support.
3763 [Jeff Dubrule <igor@pobox.org>]
3765 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
3766 bn_div_words in alpha.s.
3767 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
3769 *) Make sure the RSA OAEP test is skipped under -DRSAref because
3770 OAEP isn't supported when OpenSSL is built with RSAref.
3771 [Ulf Moeller <ulf@fitug.de>]
3773 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
3774 so they no longer are missing under -DNOPROTO.
3775 [Soren S. Jorvang <soren@t.dk>]
3778 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
3780 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
3781 doesn't work when the session is reused. Coming soon!
3784 *) Fix a security hole, that allows sessions to be reused in the wrong
3785 context thus bypassing client cert protection! All software that uses
3786 client certs and session caches in multiple contexts NEEDS PATCHING to
3787 allow session reuse! A fuller solution is in the works.
3788 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
3790 *) Some more source tree cleanups (removed obsolete files
3791 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
3792 permission on "config" script to be executable) and a fix for the INSTALL
3794 [Ulf Moeller <ulf@fitug.de>]
3796 *) Remove some legacy and erroneous uses of malloc, free instead of
3798 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
3800 *) Make rsa_oaep_test return non-zero on error.
3801 [Ulf Moeller <ulf@fitug.de>]
3803 *) Add support for native Solaris shared libraries. Configure
3804 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
3805 if someone would make that last step automatic.
3806 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
3808 *) ctx_size was not built with the right compiler during "make links". Fixed.
3811 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
3812 except NULL ciphers". This means the default cipher list will no longer
3813 enable NULL ciphers. They need to be specifically enabled e.g. with
3814 the string "DEFAULT:eNULL".
3817 *) Fix to RSA private encryption routines: if p < q then it would
3818 occasionally produce an invalid result. This will only happen with
3819 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
3822 *) Be less restrictive and allow also `perl util/perlpath.pl
3823 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
3824 because this way one can also use an interpreter named `perl5' (which is
3825 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
3826 installed as `perl').
3827 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
3829 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
3830 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
3832 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
3833 advapi32.lib to Win32 build and change the pem test comparision
3834 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
3835 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
3836 and crypto/des/ede_cbcm_enc.c.
3839 *) DES quad checksum was broken on big-endian architectures. Fixed.
3842 *) Comment out two functions in bio.h that aren't implemented. Fix up the
3843 Win32 test batch file so it (might) work again. The Win32 test batch file
3844 is horrible: I feel ill....
3847 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
3848 in e_os.h. Audit of header files to check ANSI and non ANSI
3849 sections: 10 functions were absent from non ANSI section and not exported
3850 from Windows DLLs. Fixed up libeay.num for new functions.
3853 *) Make `openssl version' output lines consistent.
3854 [Ralf S. Engelschall]
3856 *) Fix Win32 symbol export lists for BIO functions: Added
3857 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
3858 to ms/libeay{16,32}.def.
3859 [Ralf S. Engelschall]
3861 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
3862 fine under Unix and passes some trivial tests I've now added. But the
3863 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
3864 added to make sure no one expects that this stuff really works in the
3865 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
3866 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
3868 [Ralf S. Engelschall]
3870 *) Fix the generation of two part addresses in perl.
3871 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
3873 *) Add config entry for Linux on MIPS.
3874 [John Tobey <jtobey@channel1.com>]
3876 *) Make links whenever Configure is run, unless we are on Windoze.
3879 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
3880 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
3884 *) Add a useful kludge to allow package maintainers to specify compiler and
3885 other platforms details on the command line without having to patch the
3886 Configure script everytime: One now can use ``perl Configure
3887 <id>:<details>'', i.e. platform ids are allowed to have details appended
3888 to them (seperated by colons). This is treated as there would be a static
3889 pre-configured entry in Configure's %table under key <id> with value
3890 <details> and ``perl Configure <id>'' is called. So, when you want to
3891 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
3892 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
3893 now, which overrides the FreeBSD-elf entry on-the-fly.
3894 [Ralf S. Engelschall]
3896 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
3899 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
3900 on the `perl Configure ...' command line. This way one can compile
3901 OpenSSL libraries with Position Independent Code (PIC) which is needed
3902 for linking it into DSOs.
3903 [Ralf S. Engelschall]
3905 *) Remarkably, export ciphers were totally broken and no-one had noticed!
3909 *) Cleaned up the LICENSE document: The official contact for any license
3910 questions now is the OpenSSL core team under openssl-core@openssl.org.
3911 And add a paragraph about the dual-license situation to make sure people
3912 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
3913 to the OpenSSL toolkit.
3914 [Ralf S. Engelschall]
3916 *) General source tree makefile cleanups: Made `making xxx in yyy...'
3917 display consistent in the source tree and replaced `/bin/rm' by `rm'.
3918 Additonally cleaned up the `make links' target: Remove unnecessary
3919 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
3920 to speed processing and no longer clutter the display with confusing
3921 stuff. Instead only the actually done links are displayed.
3922 [Ralf S. Engelschall]
3924 *) Permit null encryption ciphersuites, used for authentication only. It used
3925 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
3926 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
3930 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
3931 signed attributes when verifying signatures (this would break them),
3932 the detached data encoding was wrong and public keys obtained using
3933 X509_get_pubkey() weren't freed.
3936 *) Add text documentation for the BUFFER functions. Also added a work around
3937 to a Win95 console bug. This was triggered by the password read stuff: the
3938 last character typed gets carried over to the next fread(). If you were
3939 generating a new cert request using 'req' for example then the last
3940 character of the passphrase would be CR which would then enter the first
3944 *) Added the new `Includes OpenSSL Cryptography Software' button as
3945 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
3946 button and can be used by applications based on OpenSSL to show the
3947 relationship to the OpenSSL project.
3948 [Ralf S. Engelschall]
3950 *) Remove confusing variables in function signatures in files
3951 ssl/ssl_lib.c and ssl/ssl.h.
3952 [Lennart Bong <lob@kulthea.stacken.kth.se>]
3954 *) Don't install bss_file.c under PREFIX/include/
3955 [Lennart Bong <lob@kulthea.stacken.kth.se>]
3957 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
3958 functions that return function pointers and has support for NT specific
3959 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
3960 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
3961 unsigned to signed types: this was killing the Win32 compile.
3964 *) Add new certificate file to stack functions,
3965 SSL_add_dir_cert_subjects_to_stack() and
3966 SSL_add_file_cert_subjects_to_stack(). These largely supplant
3967 SSL_load_client_CA_file(), and can be used to add multiple certs easily
3968 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
3969 This means that Apache-SSL and similar packages don't have to mess around
3970 to add as many CAs as they want to the preferred list.
3973 *) Experiment with doxygen documentation. Currently only partially applied to
3975 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
3976 openssl.doxy as the configuration file.
3979 *) Get rid of remaining C++-style comments which strict C compilers hate.
3980 [Ralf S. Engelschall, pointed out by Carlos Amengual]
3982 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
3983 compiled in by default: it has problems with large keys.
3986 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
3987 DH private keys and/or callback functions which directly correspond to
3988 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
3989 is needed for applications which have to configure certificates on a
3990 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
3992 For the RSA certificate situation is makes no difference, but
3993 for the DSA certificate situation this fixes the "no shared cipher"
3994 problem where the OpenSSL cipher selection procedure failed because the
3995 temporary keys were not overtaken from the context and the API provided
3996 no way to reconfigure them.
3997 The new functions now let applications reconfigure the stuff and they
3998 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
3999 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
4000 non-public-API function ssl_cert_instantiate() is used as a helper
4001 function and also to reduce code redundancy inside ssl_rsa.c.
4002 [Ralf S. Engelschall]
4004 *) Move s_server -dcert and -dkey options out of the undocumented feature
4005 area because they are useful for the DSA situation and should be
4006 recognized by the users.
4007 [Ralf S. Engelschall]
4009 *) Fix the cipher decision scheme for export ciphers: the export bits are
4010 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
4011 SSL_EXP_MASK. So, the original variable has to be used instead of the
4012 already masked variable.
4013 [Richard Levitte <levitte@stacken.kth.se>]
4015 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
4016 [Richard Levitte <levitte@stacken.kth.se>]
4018 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
4019 from `int' to `unsigned int' because it's a length and initialized by
4020 EVP_DigestFinal() which expects an `unsigned int *'.
4021 [Richard Levitte <levitte@stacken.kth.se>]
4023 *) Don't hard-code path to Perl interpreter on shebang line of Configure
4024 script. Instead use the usual Shell->Perl transition trick.
4025 [Ralf S. Engelschall]
4027 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
4028 (in addition to RSA certificates) to match the behaviour of `openssl dsa
4029 -noout -modulus' as it's already the case for `openssl rsa -noout
4030 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
4031 currently the public key is printed (a decision which was already done by
4032 `openssl dsa -modulus' in the past) which serves a similar purpose.
4033 Additionally the NO_RSA no longer completely removes the whole -modulus
4034 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
4036 [Ralf S. Engelschall]
4038 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
4039 BIO. See the source (crypto/evp/bio_ok.c) for more info.
4040 [Arne Ansper <arne@ats.cyber.ee>]
4042 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
4043 to be added. Now both 'req' and 'ca' can use new objects defined in the
4047 *) Add cool BIO that does syslog (or event log on NT).
4048 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
4050 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
4051 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
4052 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
4053 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
4056 *) Add preliminary config info for new extension code.
4059 *) Make RSA_NO_PADDING really use no padding.
4060 [Ulf Moeller <ulf@fitug.de>]
4062 *) Generate errors when private/public key check is done.
4065 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
4066 for some CRL extensions and new objects added.
4069 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
4070 key usage extension and fuller support for authority key id.
4073 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
4074 padding method for RSA, which is recommended for new applications in PKCS
4075 #1 v2.0 (RFC 2437, October 1998).
4076 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
4077 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
4078 against Bleichbacher's attack on RSA.
4079 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
4082 *) Updates to the new SSL compression code
4083 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
4085 *) Fix so that the version number in the master secret, when passed
4086 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
4087 (because the server will not accept higher), that the version number
4088 is 0x03,0x01, not 0x03,0x00
4089 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
4091 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
4092 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
4093 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
4096 *) Support for RAW extensions where an arbitrary extension can be
4097 created by including its DER encoding. See apps/openssl.cnf for
4101 *) Make sure latest Perl versions don't interpret some generated C array
4102 code as Perl array code in the crypto/err/err_genc.pl script.
4103 [Lars Weber <3weber@informatik.uni-hamburg.de>]
4105 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
4106 not many people have the assembler. Various Win32 compilation fixes and
4107 update to the INSTALL.W32 file with (hopefully) more accurate Win32
4111 *) Modify configure script 'Configure' to automatically create crypto/date.h
4112 file under Win32 and also build pem.h from pem.org. New script
4113 util/mkfiles.pl to create the MINFO file on environments that can't do a
4114 'make files': perl util/mkfiles.pl >MINFO should work.
4117 *) Major rework of DES function declarations, in the pursuit of correctness
4118 and purity. As a result, many evil casts evaporated, and some weirdness,
4119 too. You may find this causes warnings in your code. Zapping your evil
4120 casts will probably fix them. Mostly.
4123 *) Fix for a typo in asn1.h. Bug fix to object creation script
4124 obj_dat.pl. It considered a zero in an object definition to mean
4125 "end of object": none of the objects in objects.h have any zeros
4126 so it wasn't spotted.
4127 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
4129 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
4130 Masking (CBCM). In the absence of test vectors, the best I have been able
4131 to do is check that the decrypt undoes the encrypt, so far. Send me test
4132 vectors if you have them.
4135 *) Correct calculation of key length for export ciphers (too much space was
4136 allocated for null ciphers). This has not been tested!
4139 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
4140 message is now correct (it understands "crypto" and "ssl" on its
4141 command line). There is also now an "update" option. This will update
4142 the util/ssleay.num and util/libeay.num files with any new functions.
4144 perl util/mkdef.pl crypto ssl update
4145 it will update them.
4148 *) Overhauled the Perl interface (perl/*):
4149 - ported BN stuff to OpenSSL's different BN library
4150 - made the perl/ source tree CVS-aware
4151 - renamed the package from SSLeay to OpenSSL (the files still contain
4152 their history because I've copied them in the repository)
4153 - removed obsolete files (the test scripts will be replaced
4154 by better Test::Harness variants in the future)
4155 [Ralf S. Engelschall]
4157 *) First cut for a very conservative source tree cleanup:
4158 1. merge various obsolete readme texts into doc/ssleay.txt
4159 where we collect the old documents and readme texts.
4160 2. remove the first part of files where I'm already sure that we no
4161 longer need them because of three reasons: either they are just temporary
4162 files which were left by Eric or they are preserved original files where
4163 I've verified that the diff is also available in the CVS via "cvs diff
4164 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
4165 the crypto/md/ stuff).
4166 [Ralf S. Engelschall]
4168 *) More extension code. Incomplete support for subject and issuer alt
4169 name, issuer and authority key id. Change the i2v function parameters
4170 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
4171 what that's for :-) Fix to ASN1 macro which messed up
4172 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
4175 *) Preliminary support for ENUMERATED type. This is largely copied from the
4179 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
4180 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
4182 *) Make sure `make rehash' target really finds the `openssl' program.
4183 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
4185 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
4186 like to hear about it if this slows down other processors.
4189 *) Add CygWin32 platform information to Configure script.
4190 [Alan Batie <batie@aahz.jf.intel.com>]
4192 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
4193 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
4195 *) New program nseq to manipulate netscape certificate sequences
4198 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
4202 *) Fixes to BN code. Previously the default was to define BN_RECURSION
4203 but the BN code had some problems that would cause failures when
4204 doing certificate verification and some other functions.
4205 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
4207 *) Add ASN1 and PEM code to support netscape certificate sequences.
4210 *) Add ASN1 and PEM code to support netscape certificate sequences.
4213 *) Add several PKIX and private extended key usage OIDs.
4216 *) Modify the 'ca' program to handle the new extension code. Modify
4217 openssl.cnf for new extension format, add comments.
4220 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
4221 and add a sample to openssl.cnf so req -x509 now adds appropriate
4225 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
4226 error code, add initial support to X509_print() and x509 application.
4229 *) Takes a deep breath and start addding X509 V3 extension support code. Add
4230 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
4231 stuff is currently isolated and isn't even compiled yet.
4234 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
4235 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
4236 Removed the versions check from X509 routines when loading extensions:
4237 this allows certain broken certificates that don't set the version
4238 properly to be processed.
4241 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
4242 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
4243 can still be regenerated with "make depend".
4246 *) Spelling mistake in C version of CAST-128.
4247 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
4249 *) Changes to the error generation code. The perl script err-code.pl
4250 now reads in the old error codes and retains the old numbers, only
4251 adding new ones if necessary. It also only changes the .err files if new
4252 codes are added. The makefiles have been modified to only insert errors
4253 when needed (to avoid needlessly modifying header files). This is done
4254 by only inserting errors if the .err file is newer than the auto generated
4255 C file. To rebuild all the error codes from scratch (the old behaviour)
4256 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
4257 or delete all the .err files.
4260 *) CAST-128 was incorrectly implemented for short keys. The C version has
4261 been fixed, but is untested. The assembler versions are also fixed, but
4262 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
4263 to regenerate it if needed.
4264 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
4265 Hagino <itojun@kame.net>]
4267 *) File was opened incorrectly in randfile.c.
4268 [Ulf Möller <ulf@fitug.de>]
4270 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
4271 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
4272 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
4273 al: it's just almost always a UTCTime. Note this patch adds new error
4274 codes so do a "make errors" if there are problems.
4277 *) Correct Linux 1 recognition in config.
4278 [Ulf Möller <ulf@fitug.de>]
4280 *) Remove pointless MD5 hash when using DSA keys in ca.
4281 [Anonymous <nobody@replay.com>]
4283 *) Generate an error if given an empty string as a cert directory. Also
4284 generate an error if handed NULL (previously returned 0 to indicate an
4285 error, but didn't set one).
4286 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
4288 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
4291 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
4292 parameters. This was causing a warning which killed off the Win32 compile.
4295 *) Remove C++ style comments from crypto/bn/bn_local.h.
4296 [Neil Costigan <neil.costigan@celocom.com>]
4298 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
4299 based on a text string, looking up short and long names and finally
4300 "dot" format. The "dot" format stuff didn't work. Added new function
4301 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
4302 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
4303 OID is not part of the table.
4306 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
4307 X509_LOOKUP_by_alias().
4310 *) Sort openssl functions by name.
4313 *) Get the gendsa program working (hopefully) and add it to app list. Remove
4314 encryption from sample DSA keys (in case anyone is interested the password
4318 *) Make _all_ *_free functions accept a NULL pointer.
4319 [Frans Heymans <fheymans@isaserver.be>]
4321 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
4323 [Anonymous <nobody@replay.com>]
4325 *) s_server should send the CAfile as acceptable CAs, not its own cert.
4326 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
4328 *) Don't blow it for numeric -newkey arguments to apps/req.
4329 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
4331 *) Temp key "for export" tests were wrong in s3_srvr.c.
4332 [Anonymous <nobody@replay.com>]
4334 *) Add prototype for temp key callback functions
4335 SSL_CTX_set_tmp_{rsa,dh}_callback().
4338 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
4339 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
4342 *) X509_name_add_entry() freed the wrong thing after an error.
4343 [Arne Ansper <arne@ats.cyber.ee>]
4345 *) rsa_eay.c would attempt to free a NULL context.
4346 [Arne Ansper <arne@ats.cyber.ee>]
4348 *) BIO_s_socket() had a broken should_retry() on Windoze.
4349 [Arne Ansper <arne@ats.cyber.ee>]
4351 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
4352 [Arne Ansper <arne@ats.cyber.ee>]
4354 *) Make sure the already existing X509_STORE->depth variable is initialized
4355 in X509_STORE_new(), but document the fact that this variable is still
4356 unused in the certificate verification process.
4357 [Ralf S. Engelschall]
4359 *) Fix the various library and apps files to free up pkeys obtained from
4360 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
4363 *) Fix reference counting in X509_PUBKEY_get(). This makes
4364 demos/maurice/example2.c work, amongst others, probably.
4365 [Steve Henson and Ben Laurie]
4367 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
4368 `openssl' and second, the shortcut symlinks for the `openssl <command>'
4369 are no longer created. This way we have a single and consistent command
4370 line interface `openssl <command>', similar to `cvs <command>'.
4371 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
4373 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
4374 BIT STRING wrapper always have zero unused bits.
4377 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
4380 *) Make the top-level INSTALL documentation easier to understand.
4383 *) Makefiles updated to exit if an error occurs in a sub-directory
4384 make (including if user presses ^C) [Paul Sutton]
4386 *) Make Montgomery context stuff explicit in RSA data structure.
4389 *) Fix build order of pem and err to allow for generated pem.h.
4392 *) Fix renumbering bug in X509_NAME_delete_entry().
4395 *) Enhanced the err-ins.pl script so it makes the error library number
4396 global and can add a library name. This is needed for external ASN1 and
4397 other error libraries.
4400 *) Fixed sk_insert which never worked properly.
4403 *) Fix ASN1 macros so they can handle indefinite length construted
4404 EXPLICIT tags. Some non standard certificates use these: they can now
4408 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
4409 into a single doc/ssleay.txt bundle. This way the information is still
4410 preserved but no longer messes up this directory. Now it's new room for
4411 the new set of documenation files.
4412 [Ralf S. Engelschall]
4414 *) SETs were incorrectly DER encoded. This was a major pain, because they
4415 shared code with SEQUENCEs, which aren't coded the same. This means that
4416 almost everything to do with SETs or SEQUENCEs has either changed name or
4417 number of arguments.
4418 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
4420 *) Fix test data to work with the above.
4423 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
4424 was already fixed by Eric for 0.9.1 it seems.
4425 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
4427 *) Autodetect FreeBSD3.
4430 *) Fix various bugs in Configure. This affects the following platforms:
4434 unixware-2.0-pentium
4438 *) Eliminate generated files from CVS. Reorder tests to regenerate files
4439 before they are needed.
4442 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
4446 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
4448 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
4449 changed SSLeay to OpenSSL in version strings.
4450 [Ralf S. Engelschall]
4452 *) Some fixups to the top-level documents.
4455 *) Fixed the nasty bug where rsaref.h was not found under compile-time
4456 because the symlink to include/ was missing.
4457 [Ralf S. Engelschall]
4459 *) Incorporated the popular no-RSA/DSA-only patches
4460 which allow to compile a RSA-free SSLeay.
4461 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
4463 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
4464 when "ssleay" is still not found.
4465 [Ralf S. Engelschall]
4467 *) Added more platforms to Configure: Cray T3E, HPUX 11,
4468 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
4470 *) Updated the README file.
4471 [Ralf S. Engelschall]
4473 *) Added various .cvsignore files in the CVS repository subdirs
4474 to make a "cvs update" really silent.
4475 [Ralf S. Engelschall]
4477 *) Recompiled the error-definition header files and added
4478 missing symbols to the Win32 linker tables.
4479 [Ralf S. Engelschall]
4481 *) Cleaned up the top-level documents;
4482 o new files: CHANGES and LICENSE
4483 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
4484 o merged COPYRIGHT into LICENSE
4485 o removed obsolete TODO file
4486 o renamed MICROSOFT to INSTALL.W32
4487 [Ralf S. Engelschall]
4489 *) Removed dummy files from the 0.9.1b source tree:
4490 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
4491 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
4492 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
4493 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
4494 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
4495 [Ralf S. Engelschall]
4497 *) Added various platform portability fixes.
4500 *) The Genesis of the OpenSSL rpject:
4501 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
4502 Young and Tim J. Hudson created while they were working for C2Net until
4504 [The OpenSSL Project]
4507 Changes between 0.9.0b and 0.9.1b [not released]
4509 *) Updated a few CA certificates under certs/
4512 *) Changed some BIGNUM api stuff.
4515 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
4516 DGUX x86, Linux Alpha, etc.
4519 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
4520 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
4524 *) Add -strparse option to asn1pars program which parses nested
4526 [Dr Stephen Henson <shenson@bigfoot.com>]
4528 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
4531 *) DSA fix for "ca" program.
4534 *) Added "-genkey" option to "dsaparam" program.
4537 *) Added RIPE MD160 (rmd160) message digest.
4540 *) Added -a (all) option to "ssleay version" command.
4543 *) Added PLATFORM define which is the id given to Configure.
4546 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
4549 *) Extended the ASN.1 parser routines.
4552 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
4555 *) Added a BN_CTX to the BN library.
4558 *) Fixed the weak key values in DES library
4561 *) Changed API in EVP library for cipher aliases.
4564 *) Added support for RC2/64bit cipher.
4567 *) Converted the lhash library to the crypto/mem.c functions.
4570 *) Added more recognized ASN.1 object ids.
4573 *) Added more RSA padding checks for SSL/TLS.
4576 *) Added BIO proxy/filter functionality.
4579 *) Added extra_certs to SSL_CTX which can be used
4580 send extra CA certificates to the client in the CA cert chain sending
4581 process. It can be configured with SSL_CTX_add_extra_chain_cert().
4584 *) Now Fortezza is denied in the authentication phase because
4585 this is key exchange mechanism is not supported by SSLeay at all.
4588 *) Additional PKCS1 checks.
4591 *) Support the string "TLSv1" for all TLS v1 ciphers.
4594 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
4595 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
4598 *) Fixed a few memory leaks.
4601 *) Fixed various code and comment typos.
4604 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
4605 bytes sent in the client random.
4606 [Edward Bishop <ebishop@spyglass.com>]
projects / openssl.git / blob