5 Changes between 0.9.7a and 0.9.8 [xx XXX xxxx]
7 *) In crypto/ec/ec_mult.c, implement fast point multiplication with
8 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
9 will now compute a table of multiples of the generator that
10 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
11 faster (notably in the case of a single point multiplication,
13 [Nils Larsch, Bodo Moeller]
15 *) IPv6 support for certificate extensions. The various extensions
16 which use the IP:a.b.c.d can now take IPv6 addresses using the
17 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
21 *) Added an ENGINE that implements RSA by performing private key
22 exponentiations with the GMP library. The conversions to and from
23 GMP's mpz_t format aren't optimised nor are any montgomery forms
24 cached, and on x86 it appears OpenSSL's own performance has caught up.
25 However there are likely to be other architectures where GMP could
26 provide a boost. This ENGINE is not built in by default, but it can be
27 specified at Configure time and should be accompanied by the necessary
29 ./config -DOPENSSL_USE_GMP -lgmp
32 *) "openssl engine" will not display ENGINE/DSO load failure errors when
33 testing availability of engines with "-t" - the old behaviour is
34 produced by increasing the feature's verbosity with "-tt".
37 *) ECDSA routines: under certain error conditions uninitialized BN objects
38 could be freed. Solution: make sure initialization is performed early
39 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
43 *) Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
44 and DH_METHOD (eg. by ENGINE implementations) to override the normal
45 software implementations. For DSA and DH, parameter generation can
46 also be overriden by providing the appropriate method callbacks.
49 *) Change the "progress" mechanism used in key-generation and
50 primality testing to functions that take a new BN_GENCB pointer in
51 place of callback/argument pairs. The new API functions have "_ex"
52 postfixes and the older functions are reimplemented as wrappers for
53 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
54 declarations of the old functions to help (graceful) attempts to
55 migrate to the new functions. Also, the new key-generation API
56 functions operate on a caller-supplied key-structure and return
57 success/failure rather than returning a key or NULL - this is to
58 help make "keygen" another member function of RSA_METHOD etc.
60 Example for using the new callback interface:
62 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
66 BN_GENCB_set(&my_cb, my_callback, my_arg);
68 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
69 /* For the meaning of a, b in calls to my_callback(), see the
70 * documentation of the function that calls the callback.
71 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
72 * my_callback should return 1 if it wants BN_is_prime_ex()
73 * to continue, or 0 to stop.
78 *) Change the ZLIB compression method to be stateful, and make it
79 available to TLS with the number defined in
80 draft-ietf-tls-compression-04.txt.
83 *) Add the ASN.1 structures and functions for CertificatePair, which
84 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
86 CertificatePair ::= SEQUENCE {
87 forward [0] Certificate OPTIONAL,
88 reverse [1] Certificate OPTIONAL,
89 -- at least one of the pair shall be present -- }
91 Also implement the PEM functions to read and write certificate
92 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
94 This needed to be defined, mostly for the sake of the LDAP
95 attribute crossCertificatePair, but may prove useful elsewhere as
99 *) Make it possible to inhibit symlinking of shared libraries in
100 Makefile.shared, for Cygwin's sake.
103 *) Extend the BIGNUM API by creating new macros that behave like
106 void BN_set_sign(BIGNUM *a, int neg);
107 int BN_get_sign(const BIGNUM *a);
109 and avoid the need to access 'a->neg' directly in applications.
110 [Nils Larsch <nla@trustcenter.de>]
112 *) Implement fast modular reduction for pseudo-Mersenne primes
113 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
114 EC_GROUP_new_curve_GFp() will now automatically use this
116 [Nils Larsch <nla@trustcenter.de>]
118 *) Add new lock type (CRYPTO_LOCK_BN).
121 *) Change the ENGINE framework to automatically load engines
122 dynamically from specific directories unless they could be
123 found to already be built in or loaded. Move all the
124 current engines except for the cryptodev one to a new
126 The engines in engines/ are built as shared libraries if
127 the "shared" options was given to ./Configure or ./config.
128 Otherwise, they are inserted in libcrypto.a.
129 /usr/local/ssl/engines is the default directory for dynamic
130 engines, but that can be overriden at configure time through
131 the usual use of --prefix and/or --openssldir, and at run
132 time with the environment variable OPENSSL_ENGINES.
133 [Geoff Thorpe and Richard Levitte]
135 *) Add Makefile.shared, a helper makefile to build shared
136 libraries. Addapt Makefile.org.
139 *) Add version info to Win32 DLLs.
140 [Peter 'Luna' Runestig" <peter@runestig.com>]
142 *) Add new 'medium level' PKCS#12 API. Certificates and keys
143 can be added using this API to created arbitrary PKCS#12
144 files while avoiding the low level API.
146 New options to PKCS12_create(), key or cert can be NULL and
147 will then be omitted from the output file. The encryption
148 algorithm NIDs can be set to -1 for no encryption, the mac
149 iteration count can be set to 0 to omit the mac.
151 Enhance pkcs12 utility by making the -nokeys and -nocerts
152 options work when creating a PKCS#12 file. New option -nomac
153 to omit the mac, NONE can be set for an encryption algorithm.
154 New code is modified to use the enhanced PKCS12_create()
155 instead of the low level API.
158 *) Extend ASN1 encoder to support indefinite length constructed
159 encoding. This can output sequences tags and octet strings in
160 this form. Modify pk7_asn1.c to support indefinite length
161 encoding. This is experimental and needs additional code to
162 be useful, such as an ASN1 bio and some enhanced streaming
165 Extend template encode functionality so that tagging is passed
166 down to the template encoder.
169 *) Let 'openssl req' fail if an argument to '-newkey' is not
170 recognized instead of using RSA as a default.
173 *) Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
174 As these are not official, they are not included in "ALL";
175 the "ECCdraft" ciphersuite group alias can be used to select them.
176 [Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)]
178 *) Add ECDH engine support.
179 [Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)]
181 *) Add ECDH in new directory crypto/ecdh/.
182 TODO: more general interface (return x coordinate, not its hash)
183 TODO: bug: pad x with leading zeros if necessary
184 [Douglas Stebila (Sun Microsystems Laboratories)]
186 *) Let BN_rand_range() abort with an error after 100 iterations
187 without success (which indicates a broken PRNG).
190 *) Change BN_mod_sqrt() so that it verifies that the input value
191 is really the square of the return value. (Previously,
192 BN_mod_sqrt would show GIGO behaviour.)
195 *) Add named elliptic curves over binary fields from X9.62, SECG,
196 and WAP/WTLS; add OIDs that were still missing.
198 [Sheueling Chang Shantz and Douglas Stebila
199 (Sun Microsystems Laboratories)]
201 *) Extend the EC library for elliptic curves over binary fields
202 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
205 EC_GF2m_simple_method
209 EC_GROUP_new_curve_GF2m
210 EC_GROUP_set_curve_GF2m
211 EC_GROUP_get_curve_GF2m
212 EC_POINT_set_affine_coordinates_GF2m
213 EC_POINT_get_affine_coordinates_GF2m
214 EC_POINT_set_compressed_coordinates_GF2m
216 Point compression for binary fields is disabled by default for
217 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
220 As binary polynomials are represented as BIGNUMs, various members
221 of the EC_GROUP and EC_POINT data structures can be shared
222 between the implementations for prime fields and binary fields;
223 the above ..._GF2m functions (except for EX_GROUP_new_curve_GF2m)
224 are essentially identical to their ..._GFp counterparts.
225 (For simplicity, the '..._GFp' prefix has been dropped from
226 various internal method names.)
228 An internal 'field_div' method (similar to 'field_mul' and
229 'field_sqr') has been added; this is used only for binary fields.
231 [Sheueling Chang Shantz and Douglas Stebila
232 (Sun Microsystems Laboratories)]
234 *) Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
235 through methods ('mul', 'precompute_mult').
237 The generic implementations (now internally called 'ec_wNAF_mul'
238 and 'ec_wNAF_precomputed_mult') remain the default if these
239 methods are undefined.
241 [Sheueling Chang Shantz and Douglas Stebila
242 (Sun Microsystems Laboratories)]
244 *) New function EC_GROUP_get_degree, which is defined through
245 EC_METHOD. For curves over prime fields, this returns the bit
246 length of the modulus.
248 [Sheueling Chang Shantz and Douglas Stebila
249 (Sun Microsystems Laboratories)]
251 *) New functions EC_GROUP_dup, EC_POINT_dup.
252 (These simply call ..._new and ..._copy).
254 [Sheueling Chang Shantz and Douglas Stebila
255 (Sun Microsystems Laboratories)]
257 *) Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
258 Polynomials are represented as BIGNUMs (where the sign bit is not
259 used) in the following functions [macros]:
262 BN_GF2m_sub [= BN_GF2m_add]
263 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
264 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
265 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
267 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
268 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
269 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
270 BN_GF2m_cmp [= BN_ucmp]
272 (Note that only the 'mod' functions are actually for fields GF(2^m).
273 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
275 For some functions, an the irreducible polynomial defining a
276 field can be given as an 'unsigned int[]' with strictly
277 decreasing elements giving the indices of those bits that are set;
278 i.e., p[] represents the polynomial
279 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
281 p[0] > p[1] > ... > p[k] = 0.
282 This applies to the following functions:
287 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
288 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
291 BN_GF2m_mod_solve_quad_arr
295 Conversion can be performed by the following functions:
300 bntest.c has additional tests for binary polynomial arithmetic.
302 Two implementations for BN_GF2m_mod_div() are available.
303 The default algorithm simply uses BN_GF2m_mod_inv() and
304 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
305 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
306 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
308 [Sheueling Chang Shantz and Douglas Stebila
309 (Sun Microsystems Laboratories)]
311 *) Add new error code 'ERR_R_DISABLED' that can be used when some
312 functionality is disabled at compile-time.
313 [Douglas Stebila <douglas.stebila@sun.com>]
315 *) Change default behaviour of 'openssl asn1parse' so that more
316 information is visible when viewing, e.g., a certificate:
318 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
319 mode the content of non-printable OCTET STRINGs is output in a
320 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
321 avoid the appearance of a printable string.
322 [Nils Larsch <nla@trustcenter.de>]
324 *) Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
326 EC_GROUP_set_asn1_flag()
327 EC_GROUP_get_asn1_flag()
328 EC_GROUP_set_point_conversion_form()
329 EC_GROUP_get_point_conversion_form()
330 These control ASN1 encoding details:
331 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
332 has been set to OPENSSL_EC_NAMED_CURVE.
333 - Points are encoded in uncompressed form by default; options for
334 asn1_for are as for point2oct, namely
335 POINT_CONVERSION_COMPRESSED
336 POINT_CONVERSION_UNCOMPRESSED
337 POINT_CONVERSION_HYBRID
339 Also add 'seed' and 'seed_len' members to EC_GROUP with access
343 EC_GROUP_get_seed_len()
344 This is used only for ASN1 purposes (so far).
345 [Nils Larsch <nla@trustcenter.de>]
347 *) Add 'field_type' member to EC_METHOD, which holds the NID
348 of the appropriate field type OID. The new function
349 EC_METHOD_get_field_type() returns this value.
350 [Nils Larsch <nla@trustcenter.de>]
357 providing useful interfaces to EC_POINT_point2oct() and
358 EC_POINT_oct2point().
359 [Nils Larsch <nla@trustcenter.de>]
361 *) Change internals of the EC library so that the functions
362 EC_GROUP_set_generator()
363 EC_GROUP_get_generator()
365 EC_GROUP_get_cofactor()
366 are implemented directly in crypto/ec/ec_lib.c and not dispatched
367 to methods, which would lead to unnecessary code duplication when
368 adding different types of curves.
369 [Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller]
371 *) Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
372 arithmetic, and such that modified wNAFs are generated
373 (which avoid length expansion in many cases).
376 *) Add a function EC_GROUP_check_discriminant() (defined via
377 EC_METHOD) that verifies that the curve discriminant is non-zero.
379 Add a function EC_GROUP_check() that makes some sanity tests
380 on a EC_GROUP, its generator and order. This includes
381 EC_GROUP_check_discriminant().
382 [Nils Larsch <nla@trustcenter.de>]
384 *) Add ECDSA in new directory crypto/ecdsa/.
386 Add applications 'openssl ecparam' and 'openssl ecdsa'
387 (these are based on 'openssl dsaparam' and 'openssl dsa').
389 ECDSA support is also included in various other files across the
390 library. Most notably,
391 - 'openssl req' now has a '-newkey ecdsa:file' option;
392 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
393 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
394 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
395 them suitable for ECDSA where domain parameters must be
396 extracted before the specific public key;
397 - ECDSA engine support has been added.
398 [Nils Larsch <nla@trustcenter.de>]
400 *) Include some named elliptic curves, and add OIDs from X9.62,
401 SECG, and WAP/WTLS. Each curve can be obtained from the new
403 EC_GROUP_new_by_nid(),
404 and the list of available named curves can be obtained with
405 EC_get_builtin_curves().
406 Also add a 'curve_name' member to EC_GROUP objects, which can be
410 [Nils Larsch <nla@trustcenter.de, Bodo Moeller]
412 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
413 was actually never needed) and in BN_mul(). The removal in BN_mul()
414 required a small change in bn_mul_part_recursive() and the addition
415 of the functions bn_cmp_part_words(), bn_sub_part_words() and
416 bn_add_part_words(), which do the same thing as bn_cmp_words(),
417 bn_sub_words() and bn_add_words() except they take arrays with
421 Changes between 0.9.7 and 0.9.7a [XX xxx 2003]
423 *) Allow an application to disable the automatic SSL chain building.
424 Before this a rather primitive chain build was always performed in
425 ssl3_output_cert_chain(): an application had no way to send the
426 correct chain if the automatic operation produced an incorrect result.
428 Now the chain builder is disabled if either:
430 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
432 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
434 The reasoning behind this is that an application would not want the
435 auto chain building to take place if extra chain certificates are
436 present and it might also want a means of sending no additional
437 certificates (for example the chain has two certificates and the
441 *) Add the possibility to build without the ENGINE framework.
442 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
444 *) Under Win32 gmtime() can return NULL: check return value in
445 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
448 *) DSA routines: under certain error conditions uninitialized BN objects
449 could be freed. Solution: make sure initialization is performed early
450 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
451 Nils Larsch <nla@trustcenter.de> via PR#459)
454 *) Another fix for SSLv2 session ID handling: the session ID was incorrectly
455 checked on reconnect on the client side, therefore session resumption
456 could still fail with a "ssl session id is different" error. This
457 behaviour is masked when SSL_OP_ALL is used due to
458 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
459 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
463 *) IA-32 assembler support enhancements: unified ELF targets, support
464 for SCO/Caldera platforms, fix for Cygwin shared build.
467 *) Add support for FreeBSD on sparc64. As a consequence, support for
468 FreeBSD on non-x86 processors is separate from x86 processors on
469 the config script, much like the NetBSD support.
470 [Richard Levitte & Kris Kennaway <kris@obsecurity.org>]
472 Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
474 *) Fix session ID handling in SSLv2 client code: the SERVER FINISHED
475 code (06) was taken as the first octet of the session ID and the last
476 octet was ignored consequently. As a result SSLv2 client side session
477 caching could not have worked due to the session ID mismatch between
479 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
483 *) Change the declaration of needed Kerberos libraries to use EX_LIBS
484 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
488 *) The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
489 seems that in spite of existing for more than a year, many application
490 author have done nothing to provide the necessary callbacks, which
491 means that this particular engine will not work properly anywhere.
492 This is a very unfortunate situation which forces us, in the name
493 of usability, to give the hw_ncipher.c a static lock, which is part
495 NOTE: This is for the 0.9.7 series ONLY. This hack will never
496 appear in 0.9.8 or later. We EXPECT application authors to have
497 dealt properly with this when 0.9.8 is released (unless we actually
498 make such changes in the libcrypto locking code that changes will
499 have to be made anyway).
502 *) In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
503 octets have been read, EOF or an error occurs. Without this change
504 some truncated ASN1 structures will not produce an error.
507 *) Disable Heimdal support, since it hasn't been fully implemented.
508 Still give the possibility to force the use of Heimdal, but with
509 warnings and a request that patches get sent to openssl-dev.
512 *) Add the VC-CE target, introduce the WINCE sysname, and add
513 INSTALL.WCE and appropriate conditionals to make it build.
514 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
516 *) Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
517 cygssl-x.y.z.dll, where x, y and z are the major, minor and
518 edit numbers of the version.
519 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
521 *) Introduce safe string copy and catenation functions
522 (BUF_strlcpy() and BUF_strlcat()).
523 [Ben Laurie (CHATS) and Richard Levitte]
525 *) Avoid using fixed-size buffers for one-line DNs.
528 *) Add BUF_MEM_grow_clean() to avoid information leakage when
529 resizing buffers containing secrets, and use where appropriate.
532 *) Avoid using fixed size buffers for configuration file location.
535 *) Avoid filename truncation for various CA files.
538 *) Use sizeof in preference to magic numbers.
541 *) Avoid filename truncation in cert requests.
544 *) Add assertions to check for (supposedly impossible) buffer
548 *) Don't cache truncated DNS entries in the local cache (this could
549 potentially lead to a spoofing attack).
552 *) Fix various buffers to be large enough for hex/decimal
553 representations in a platform independent manner.
556 *) Add CRYPTO_realloc_clean() to avoid information leakage when
557 resizing buffers containing secrets, and use where appropriate.
560 *) Add BIO_indent() to avoid much slightly worrying code to do
564 *) Convert sprintf()/BIO_puts() to BIO_printf().
567 *) buffer_gets() could terminate with the buffer only half
571 *) Add assertions to prevent user-supplied crypto functions from
572 overflowing internal buffers by having large block sizes, etc.
575 *) New OPENSSL_assert() macro (similar to assert(), but enabled
579 *) Eliminate unused copy of key in RC4.
582 *) Eliminate unused and incorrectly sized buffers for IV in pem.h.
585 *) Fix off-by-one error in EGD path.
588 *) If RANDFILE path is too long, ignore instead of truncating.
591 *) Eliminate unused and incorrectly sized X.509 structure
595 *) Eliminate unused and dangerous function knumber().
598 *) Eliminate unused and dangerous structure, KSSL_ERR.
601 *) Protect against overlong session ID context length in an encoded
602 session object. Since these are local, this does not appear to be
606 *) Change from security patch (see 0.9.6e below) that did not affect
607 the 0.9.6 release series:
609 Remote buffer overflow in SSL3 protocol - an attacker could
610 supply an oversized master key in Kerberos-enabled versions.
614 *) Change the SSL kerb5 codes to match RFC 2712.
617 *) Make -nameopt work fully for req and add -reqopt switch.
618 [Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson]
620 *) The "block size" for block ciphers in CFB and OFB mode should be 1.
621 [Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>]
623 *) Make sure tests can be performed even if the corresponding algorithms
624 have been removed entirely. This was also the last step to make
625 OpenSSL compilable with DJGPP under all reasonable conditions.
626 [Richard Levitte, Doug Kaufman <dkaufman@rahul.net>]
628 *) Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
629 to allow version independent disabling of normally unselected ciphers,
630 which may be activated as a side-effect of selecting a single cipher.
632 (E.g., cipher list string "RSA" enables ciphersuites that are left
633 out of "ALL" because they do not provide symmetric encryption.
634 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
635 [Lutz Jaenicke, Bodo Moeller]
637 *) Add appropriate support for separate platform-dependent build
638 directories. The recommended way to make a platform-dependent
639 build directory is the following (tested on Linux), maybe with
642 # Place yourself outside of the OpenSSL source tree. In
643 # this example, the environment variable OPENSSL_SOURCE
644 # is assumed to contain the absolute OpenSSL source directory.
645 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
646 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
647 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
648 mkdir -p `dirname $F`
649 ln -s $OPENSSL_SOURCE/$F $F
652 To be absolutely sure not to disturb the source tree, a "make clean"
653 is a good thing. If it isn't successfull, don't worry about it,
654 it probably means the source directory is very clean.
657 *) Make sure any ENGINE control commands make local copies of string
658 pointers passed to them whenever necessary. Otherwise it is possible
659 the caller may have overwritten (or deallocated) the original string
660 data when a later ENGINE operation tries to use the stored values.
661 [Götz Babin-Ebell <babinebell@trustcenter.de>]
663 *) Improve diagnostics in file reading and command-line digests.
664 [Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>]
666 *) Add AES modes CFB and OFB to the object database. Correct an
667 error in AES-CFB decryption.
670 *) Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
671 allows existing EVP_CIPHER_CTX structures to be reused after
672 calling EVP_*Final(). This behaviour is used by encryption
673 BIOs and some applications. This has the side effect that
674 applications must explicitly clean up cipher contexts with
675 EVP_CIPHER_CTX_cleanup() or they will leak memory.
678 *) Check the values of dna and dnb in bn_mul_recursive before calling
679 bn_mul_comba (a non zero value means the a or b arrays do not contain
680 n2 elements) and fallback to bn_mul_normal if either is not zero.
683 *) Fix escaping of non-ASCII characters when using the -subj option
684 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
687 *) Make object definitions compliant to LDAP (RFC2256): SN is the short
688 form for "surname", serialNumber has no short form.
689 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
690 therefore remove "mail" short name for "internet 7".
691 The OID for unique identifiers in X509 certificates is
692 x500UniqueIdentifier, not uniqueIdentifier.
693 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
696 *) Add an "init" command to the ENGINE config module and auto initialize
697 ENGINEs. Without any "init" command the ENGINE will be initialized
698 after all ctrl commands have been executed on it. If init=1 the
699 ENGINE is initailized at that point (ctrls before that point are run
700 on the uninitialized ENGINE and after on the initialized one). If
701 init=0 then the ENGINE will not be iniatialized at all.
704 *) Fix the 'app_verify_callback' interface so that the user-defined
705 argument is actually passed to the callback: In the
706 SSL_CTX_set_cert_verify_callback() prototype, the callback
707 declaration has been changed from
710 int (*cb)(X509_STORE_CTX *,void *);
711 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
712 i=s->ctx->app_verify_callback(&ctx)
713 has been changed into
714 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
716 To update applications using SSL_CTX_set_cert_verify_callback(),
717 a dummy argument can be added to their callback functions.
718 [D. K. Smetters <smetters@parc.xerox.com>]
720 *) Added the '4758cca' ENGINE to support IBM 4758 cards.
721 [Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe]
723 *) Add and OPENSSL_LOAD_CONF define which will cause
724 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
725 This allows older applications to transparently support certain
726 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
727 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
728 load the config file and OPENSSL_add_all_algorithms_conf() which will
729 always load it have also been added.
732 *) Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
733 Adjust NIDs and EVP layer.
734 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
736 *) Config modules support in openssl utility.
738 Most commands now load modules from the config file,
739 though in a few (such as version) this isn't done
740 because it couldn't be used for anything.
742 In the case of ca and req the config file used is
743 the same as the utility itself: that is the -config
744 command line option can be used to specify an
748 *) Move default behaviour from OPENSSL_config(). If appname is NULL
749 use "openssl_conf" if filename is NULL use default openssl config file.
752 *) Add an argument to OPENSSL_config() to allow the use of an alternative
753 config section name. Add a new flag to tolerate a missing config file
754 and move code to CONF_modules_load_file().
757 *) Support for crypto accelerator cards from Accelerated Encryption
758 Processing, www.aep.ie. (Use engine 'aep')
759 The support was copied from 0.9.6c [engine] and adapted/corrected
760 to work with the new engine framework.
761 [AEP Inc. and Richard Levitte]
763 *) Support for SureWare crypto accelerator cards from Baltimore
764 Technologies. (Use engine 'sureware')
765 The support was copied from 0.9.6c [engine] and adapted
766 to work with the new engine framework.
769 *) Have the CHIL engine fork-safe (as defined by nCipher) and actually
770 make the newer ENGINE framework commands for the CHIL engine work.
771 [Toomas Kiisk <vix@cyber.ee> and Richard Levitte]
773 *) Make it possible to produce shared libraries on ReliantUNIX.
774 [Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte]
776 *) Add the configuration target debug-linux-ppro.
777 Make 'openssl rsa' use the general key loading routines
778 implemented in apps.c, and make those routines able to
779 handle the key format FORMAT_NETSCAPE and the variant
781 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
783 *) Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
784 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
786 *) Add -keyform to rsautl, and document -engine.
787 [Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>]
789 *) Change BIO_new_file (crypto/bio/bss_file.c) to use new
790 BIO_R_NO_SUCH_FILE error code rather than the generic
791 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
796 ERR_peek_last_error_line
797 ERR_peek_last_error_line_data.
801 ERR_peek_error_line_data,
802 but report on the latest error recorded rather than the first one
803 still in the error queue.
804 [Ben Laurie, Bodo Moeller]
806 *) default_algorithms option in ENGINE config module. This allows things
808 default_algorithms = ALL
809 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
812 *) Prelminary ENGINE config module.
815 *) New experimental application configuration code.
818 *) Change the AES code to follow the same name structure as all other
819 symmetric ciphers, and behave the same way. Move everything to
820 the directory crypto/aes, thereby obsoleting crypto/rijndael.
821 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
823 *) SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
824 [Ben Laurie and Theo de Raadt]
826 *) Add option to output public keys in req command.
827 [Massimiliano Pala madwolf@openca.org]
829 *) Use wNAFs in EC_POINTs_mul() for improved efficiency
830 (up to about 10% better than before for P-192 and P-224).
833 *) New functions/macros
835 SSL_CTX_set_msg_callback(ctx, cb)
836 SSL_CTX_set_msg_callback_arg(ctx, arg)
837 SSL_set_msg_callback(ssl, cb)
838 SSL_set_msg_callback_arg(ssl, arg)
840 to request calling a callback function
842 void cb(int write_p, int version, int content_type,
843 const void *buf, size_t len, SSL *ssl, void *arg)
845 whenever a protocol message has been completely received
846 (write_p == 0) or sent (write_p == 1). Here 'version' is the
847 protocol version according to which the SSL library interprets
848 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
849 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
850 the content type as defined in the SSL 3.0/TLS 1.0 protocol
851 specification (change_cipher_spec(20), alert(21), handshake(22)).
852 'buf' and 'len' point to the actual message, 'ssl' to the
853 SSL object, and 'arg' is the application-defined value set by
854 SSL[_CTX]_set_msg_callback_arg().
856 'openssl s_client' and 'openssl s_server' have new '-msg' options
857 to enable a callback that displays all protocol messages.
860 *) Change the shared library support so shared libraries are built as
861 soon as the corresponding static library is finished, and thereby get
862 openssl and the test programs linked against the shared library.
863 This still only happens when the keyword "shard" has been given to
864 the configuration scripts.
866 NOTE: shared library support is still an experimental thing, and
867 backward binary compatibility is still not guaranteed.
868 ["Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte]
870 *) Add support for Subject Information Access extension.
871 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
873 *) Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
874 additional bytes when new memory had to be allocated, not just
875 when reusing an existing buffer.
878 *) New command line and configuration option 'utf8' for the req command.
879 This allows field values to be specified as UTF8 strings.
882 *) Add -multi and -mr options to "openssl speed" - giving multiple parallel
883 runs for the former and machine-readable output for the latter.
886 *) Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
887 of the e-mail address in the DN (i.e., it will go into a certificate
888 extension only). The new configuration file option 'email_in_dn = no'
890 [Massimiliano Pala madwolf@openca.org]
892 *) Change all functions with names starting with des_ to be starting
893 with DES_ instead. Add wrappers that are compatible with libdes,
894 but are named _ossl_old_des_*. Finally, add macros that map the
895 des_* symbols to the corresponding _ossl_old_des_* if libdes
896 compatibility is desired. If OpenSSL 0.9.6c compatibility is
897 desired, the des_* symbols will be mapped to DES_*, with one
900 Since we provide two compatibility mappings, the user needs to
901 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
902 compatibility is desired. The default (i.e., when that macro
903 isn't defined) is OpenSSL 0.9.6c compatibility.
905 There are also macros that enable and disable the support of old
906 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
907 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
908 are defined, the default will apply: to support the old des routines.
910 In either case, one must include openssl/des.h to get the correct
911 definitions. Do not try to just include openssl/des_old.h, that
914 NOTE: This is a major break of an old API into a new one. Software
915 authors are encouraged to switch to the DES_ style functions. Some
916 time in the future, des_old.h and the libdes compatibility functions
917 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
918 default), and then completely removed.
921 *) Test for certificates which contain unsupported critical extensions.
922 If such a certificate is found during a verify operation it is
923 rejected by default: this behaviour can be overridden by either
924 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
925 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
926 X509_supported_extension() has also been added which returns 1 if a
927 particular extension is supported.
930 *) Modify the behaviour of EVP cipher functions in similar way to digests
931 to retain compatibility with existing code.
934 *) Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
935 compatibility with existing code. In particular the 'ctx' parameter does
936 not have to be to be initialized before the call to EVP_DigestInit() and
937 it is tidied up after a call to EVP_DigestFinal(). New function
938 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
939 EVP_MD_CTX_copy() changed to not require the destination to be
940 initialized valid and new function EVP_MD_CTX_copy_ex() added which
941 requires the destination to be valid.
943 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
944 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
947 *) Change ssl3_get_message (ssl/s3_both.c) and the functions using it
948 so that complete 'Handshake' protocol structures are kept in memory
949 instead of overwriting 'msg_type' and 'length' with 'body' data.
952 *) Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
953 [Massimo Santin via Richard Levitte]
955 *) Major restructuring to the underlying ENGINE code. This includes
956 reduction of linker bloat, separation of pure "ENGINE" manipulation
957 (initialisation, etc) from functionality dealing with implementations
958 of specific crypto iterfaces. This change also introduces integrated
959 support for symmetric ciphers and digest implementations - so ENGINEs
960 can now accelerate these by providing EVP_CIPHER and EVP_MD
961 implementations of their own. This is detailed in crypto/engine/README
962 as it couldn't be adequately described here. However, there are a few
963 API changes worth noting - some RSA, DSA, DH, and RAND functions that
964 were changed in the original introduction of ENGINE code have now
965 reverted back - the hooking from this code to ENGINE is now a good
966 deal more passive and at run-time, operations deal directly with
967 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
968 dereferencing through an ENGINE pointer any more. Also, the ENGINE
969 functions dealing with BN_MOD_EXP[_CRT] handlers have been removed -
970 they were not being used by the framework as there is no concept of a
971 BIGNUM_METHOD and they could not be generalised to the new
972 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
973 ENGINE_cpy() has been removed as it cannot be consistently defined in
977 *) Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
980 *) Change mkdef.pl to sort symbols that get the same entry number,
981 and make sure the automatically generated functions ERR_load_*
982 become part of libeay.num as well.
985 *) New function SSL_renegotiate_pending(). This returns true once
986 renegotiation has been requested (either SSL_renegotiate() call
987 or HelloRequest/ClientHello receveived from the peer) and becomes
988 false once a handshake has been completed.
989 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
990 sends a HelloRequest, but does not ensure that a handshake takes
991 place. SSL_renegotiate_pending() is useful for checking if the
992 client has followed the request.)
995 *) New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
996 By default, clients may request session resumption even during
997 renegotiation (if session ID contexts permit); with this option,
998 session resumption is possible only in the first handshake.
1000 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
1001 more bits available for options that should not be part of
1002 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
1005 *) Add some demos for certificate and certificate request creation.
1008 *) Make maximum certificate chain size accepted from the peer application
1009 settable (SSL*_get/set_max_cert_list()), as proposed by
1010 "Douglas E. Engert" <deengert@anl.gov>.
1013 *) Add support for shared libraries for Unixware-7
1014 (Boyd Lynn Gerber <gerberb@zenez.com>).
1017 *) Add a "destroy" handler to ENGINEs that allows structural cleanup to
1018 be done prior to destruction. Use this to unload error strings from
1019 ENGINEs that load their own error strings. NB: This adds two new API
1020 functions to "get" and "set" this destroy handler in an ENGINE.
1023 *) Alter all existing ENGINE implementations (except "openssl" and
1024 "openbsd") to dynamically instantiate their own error strings. This
1025 makes them more flexible to be built both as statically-linked ENGINEs
1026 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
1027 Also, add stub code to each that makes building them as self-contained
1028 shared-libraries easier (see README.ENGINE).
1031 *) Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
1032 implementations into applications that are completely implemented in
1033 self-contained shared-libraries. The "dynamic" ENGINE exposes control
1034 commands that can be used to configure what shared-library to load and
1035 to control aspects of the way it is handled. Also, made an update to
1036 the README.ENGINE file that brings its information up-to-date and
1037 provides some information and instructions on the "dynamic" ENGINE
1038 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
1041 *) Make it possible to unload ranges of ERR strings with a new
1042 "ERR_unload_strings" function.
1045 *) Add a copy() function to EVP_MD.
1048 *) Make EVP_MD routines take a context pointer instead of just the
1049 md_data void pointer.
1052 *) Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
1053 that the digest can only process a single chunk of data
1054 (typically because it is provided by a piece of
1055 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
1056 is only going to provide a single chunk of data, and hence the
1057 framework needn't accumulate the data for oneshot drivers.
1060 *) As with "ERR", make it possible to replace the underlying "ex_data"
1061 functions. This change also alters the storage and management of global
1062 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
1063 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
1064 index counters. The API functions that use this state have been changed
1065 to take a "class_index" rather than pointers to the class's local STACK
1066 and counter, and there is now an API function to dynamically create new
1067 classes. This centralisation allows us to (a) plug a lot of the
1068 thread-safety problems that existed, and (b) makes it possible to clean
1069 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
1070 such data would previously have always leaked in application code and
1071 workarounds were in place to make the memory debugging turn a blind eye
1072 to it. Application code that doesn't use this new function will still
1073 leak as before, but their memory debugging output will announce it now
1074 rather than letting it slide.
1076 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
1077 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
1078 has a return value to indicate success or failure.
1081 *) Make it possible to replace the underlying "ERR" functions such that the
1082 global state (2 LHASH tables and 2 locks) is only used by the "default"
1083 implementation. This change also adds two functions to "get" and "set"
1084 the implementation prior to it being automatically set the first time
1085 any other ERR function takes place. Ie. an application can call "get",
1086 pass the return value to a module it has just loaded, and that module
1087 can call its own "set" function using that value. This means the
1088 module's "ERR" operations will use (and modify) the error state in the
1089 application and not in its own statically linked copy of OpenSSL code.
1092 *) Give DH, DSA, and RSA types their own "**_up_ref()" function to increment
1093 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
1094 the operation, and provides a more encapsulated way for external code
1095 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
1096 to use these functions rather than manually incrementing the counts.
1098 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
1101 *) Add EVP test program.
1104 *) Add symmetric cipher support to ENGINE. Expect the API to change!
1107 *) New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
1108 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
1109 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
1110 These allow a CRL to be built without having to access X509_CRL fields
1111 directly. Modify 'ca' application to use new functions.
1114 *) Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
1115 bug workarounds. Rollback attack detection is a security feature.
1116 The problem will only arise on OpenSSL servers when TLSv1 is not
1117 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
1118 Software authors not wanting to support TLSv1 will have special reasons
1119 for their choice and can explicitly enable this option.
1120 [Bodo Moeller, Lutz Jaenicke]
1122 *) Rationalise EVP so it can be extended: don't include a union of
1123 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
1124 (similar to those existing for EVP_CIPHER_CTX).
1129 EVP_MD_CTX_init(&md); /* new function call */
1130 EVP_DigestInit(&md, EVP_sha1());
1131 EVP_DigestUpdate(&md, in, len);
1132 EVP_DigestFinal(&md, out, NULL);
1133 EVP_MD_CTX_cleanup(&md); /* new function call */
1137 *) Make DES key schedule conform to the usual scheme, as well as
1138 correcting its structure. This means that calls to DES functions
1139 now have to pass a pointer to a des_key_schedule instead of a
1140 plain des_key_schedule (which was actually always a pointer
1143 des_key_schedule ks;
1145 des_set_key_checked(..., &ks);
1146 des_ncbc_encrypt(..., &ks, ...);
1148 (Note that a later change renames 'des_...' into 'DES_...'.)
1151 *) Initial reduction of linker bloat: the use of some functions, such as
1152 PEM causes large amounts of unused functions to be linked in due to
1153 poor organisation. For example pem_all.c contains every PEM function
1154 which has a knock on effect of linking in large amounts of (unused)
1155 ASN1 code. Grouping together similar functions and splitting unrelated
1156 functions prevents this.
1159 *) Cleanup of EVP macros.
1162 *) Change historical references to {NID,SN,LN}_des_ede and ede3 to add the
1163 correct _ecb suffix.
1166 *) Add initial OCSP responder support to ocsp application. The
1167 revocation information is handled using the text based index
1168 use by the ca application. The responder can either handle
1169 requests generated internally, supplied in files (for example
1170 via a CGI script) or using an internal minimal server.
1173 *) Add configuration choices to get zlib compression for TLS.
1176 *) Changes to Kerberos SSL for RFC 2712 compliance:
1177 1. Implemented real KerberosWrapper, instead of just using
1178 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
1179 2. Implemented optional authenticator field of KerberosWrapper.
1181 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
1182 and authenticator structs; see crypto/krb5/.
1184 Generalized Kerberos calls to support multiple Kerberos libraries.
1185 [Vern Staats <staatsvr@asc.hpc.mil>,
1186 Jeffrey Altman <jaltman@columbia.edu>
1187 via Richard Levitte]
1189 *) Cause 'openssl speed' to use fully hard-coded DSA keys as it
1190 already does with RSA. testdsa.h now has 'priv_key/pub_key'
1191 values for each of the key sizes rather than having just
1192 parameters (and 'speed' generating keys each time).
1195 *) Speed up EVP routines.
1198 type 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
1199 des-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
1200 des-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
1201 des-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
1203 des-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
1204 des-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
1205 des-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
1208 des-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
1210 des-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
1213 *) Added the OS2-EMX target.
1214 ["Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte]
1216 *) Rewrite apps to use NCONF routines instead of the old CONF. New functions
1217 to support NCONF routines in extension code. New function CONF_set_nconf()
1218 to allow functions which take an NCONF to also handle the old LHASH
1219 structure: this means that the old CONF compatible routines can be
1220 retained (in particular wrt extensions) without having to duplicate the
1221 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
1224 *) Enhance the general user interface with mechanisms for inner control
1225 and with possibilities to have yes/no kind of prompts.
1228 *) Change all calls to low level digest routines in the library and
1229 applications to use EVP. Add missing calls to HMAC_cleanup() and
1230 don't assume HMAC_CTX can be copied using memcpy().
1231 [Verdon Walker <VWalker@novell.com>, Steve Henson]
1233 *) Add the possibility to control engines through control names but with
1234 arbitrary arguments instead of just a string.
1235 Change the key loaders to take a UI_METHOD instead of a callback
1236 function pointer. NOTE: this breaks binary compatibility with earlier
1237 versions of OpenSSL [engine].
1238 Adapt the nCipher code for these new conditions and add a card insertion
1242 *) Enhance the general user interface with mechanisms to better support
1243 dialog box interfaces, application-defined prompts, the possibility
1244 to use defaults (for example default passwords from somewhere else)
1245 and interrupts/cancellations.
1248 *) Tidy up PKCS#12 attribute handling. Add support for the CSP name
1249 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
1252 *) Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
1253 tidy up some unnecessarily weird code in 'sk_new()').
1254 [Geoff, reported by Diego Tartara <dtartara@novamens.com>]
1256 *) Change the key loading routines for ENGINEs to use the same kind
1257 callback (pem_password_cb) as all other routines that need this
1261 *) Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
1262 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
1263 than this minimum value is recommended.
1266 *) New random seeder for OpenVMS, using the system process statistics
1267 that are easily reachable.
1270 *) Windows apparently can't transparently handle global
1271 variables defined in DLLs. Initialisations such as:
1273 const ASN1_ITEM *it = &ASN1_INTEGER_it;
1275 wont compile. This is used by the any applications that need to
1276 declare their own ASN1 modules. This was fixed by adding the option
1277 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
1278 needed for static libraries under Win32.
1281 *) New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
1282 setting of purpose and trust fields. New X509_STORE trust and
1283 purpose functions and tidy up setting in other SSL functions.
1286 *) Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
1287 structure. These are inherited by X509_STORE_CTX when it is
1288 initialised. This allows various defaults to be set in the
1289 X509_STORE structure (such as flags for CRL checking and custom
1290 purpose or trust settings) for functions which only use X509_STORE_CTX
1291 internally such as S/MIME.
1293 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
1294 trust settings if they are not set in X509_STORE. This allows X509_STORE
1295 purposes and trust (in S/MIME for example) to override any set by default.
1297 Add command line options for CRL checking to smime, s_client and s_server
1301 *) Initial CRL based revocation checking. If the CRL checking flag(s)
1302 are set then the CRL is looked up in the X509_STORE structure and
1303 its validity and signature checked, then if the certificate is found
1304 in the CRL the verify fails with a revoked error.
1306 Various new CRL related callbacks added to X509_STORE_CTX structure.
1308 Command line options added to 'verify' application to support this.
1310 This needs some additional work, such as being able to handle multiple
1311 CRLs with different times, extension based lookup (rather than just
1312 by subject name) and ultimately more complete V2 CRL extension
1316 *) Add a general user interface API (crypto/ui/). This is designed
1317 to replace things like des_read_password and friends (backward
1318 compatibility functions using this new API are provided).
1319 The purpose is to remove prompting functions from the DES code
1320 section as well as provide for prompting through dialog boxes in
1321 a window system and the like.
1324 *) Add "ex_data" support to ENGINE so implementations can add state at a
1325 per-structure level rather than having to store it globally.
1328 *) Make it possible for ENGINE structures to be copied when retrieved by
1329 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
1330 This causes the "original" ENGINE structure to act like a template,
1331 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
1332 operational state can be localised to each ENGINE structure, despite the
1333 fact they all share the same "methods". New ENGINE structures returned in
1334 this case have no functional references and the return value is the single
1335 structural reference. This matches the single structural reference returned
1336 by ENGINE_by_id() normally, when it is incremented on the pre-existing
1340 *) Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
1341 needs to match any other type at all we need to manually clear the
1345 *) Changes to the "openssl engine" utility to include;
1346 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
1347 about an ENGINE's available control commands.
1348 - executing control commands from command line arguments using the
1349 '-pre' and '-post' switches. '-post' is only used if '-t' is
1350 specified and the ENGINE is successfully initialised. The syntax for
1351 the individual commands are colon-separated, for example;
1352 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
1355 *) New dynamic control command support for ENGINEs. ENGINEs can now
1356 declare their own commands (numbers), names (strings), descriptions,
1357 and input types for run-time discovery by calling applications. A
1358 subset of these commands are implicitly classed as "executable"
1359 depending on their input type, and only these can be invoked through
1360 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
1361 can be based on user input, config files, etc). The distinction is
1362 that "executable" commands cannot return anything other than a boolean
1363 result and can only support numeric or string input, whereas some
1364 discoverable commands may only be for direct use through
1365 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
1366 pointers, or other custom uses. The "executable" commands are to
1367 support parameterisations of ENGINE behaviour that can be
1368 unambiguously defined by ENGINEs and used consistently across any
1369 OpenSSL-based application. Commands have been added to all the
1370 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
1371 control over shared-library paths without source code alterations.
1374 *) Changed all ENGINE implementations to dynamically allocate their
1375 ENGINEs rather than declaring them statically. Apart from this being
1376 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
1377 this also allows the implementations to compile without using the
1378 internal engine_int.h header.
1381 *) Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
1382 'const' value. Any code that should be able to modify a RAND_METHOD
1383 should already have non-const pointers to it (ie. they should only
1384 modify their own ones).
1387 *) Made a variety of little tweaks to the ENGINE code.
1388 - "atalla" and "ubsec" string definitions were moved from header files
1389 to C code. "nuron" string definitions were placed in variables
1390 rather than hard-coded - allowing parameterisation of these values
1391 later on via ctrl() commands.
1392 - Removed unused "#if 0"'d code.
1393 - Fixed engine list iteration code so it uses ENGINE_free() to release
1394 structural references.
1395 - Constified the RAND_METHOD element of ENGINE structures.
1396 - Constified various get/set functions as appropriate and added
1397 missing functions (including a catch-all ENGINE_cpy that duplicates
1398 all ENGINE values onto a new ENGINE except reference counts/state).
1399 - Removed NULL parameter checks in get/set functions. Setting a method
1400 or function to NULL is a way of cancelling out a previously set
1401 value. Passing a NULL ENGINE parameter is just plain stupid anyway
1402 and doesn't justify the extra error symbols and code.
1403 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
1404 flags from engine_int.h to engine.h.
1405 - Changed prototypes for ENGINE handler functions (init(), finish(),
1406 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
1409 *) Implement binary inversion algorithm for BN_mod_inverse in addition
1410 to the algorithm using long division. The binary algorithm can be
1411 used only if the modulus is odd. On 32-bit systems, it is faster
1412 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
1413 roughly 5-15% for 256-bit moduli), so we use it only for moduli
1414 up to 450 bits. In 64-bit environments, the binary algorithm
1415 appears to be advantageous for much longer moduli; here we use it
1416 for moduli up to 2048 bits.
1419 *) Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
1420 could not support the combine flag in choice fields.
1423 *) Add a 'copy_extensions' option to the 'ca' utility. This copies
1424 extensions from a certificate request to the certificate.
1427 *) Allow multiple 'certopt' and 'nameopt' options to be separated
1428 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
1429 file: this allows the display of the certificate about to be
1430 signed to be customised, to allow certain fields to be included
1431 or excluded and extension details. The old system didn't display
1432 multicharacter strings properly, omitted fields not in the policy
1433 and couldn't display additional details such as extensions.
1436 *) Function EC_POINTs_mul for multiple scalar multiplication
1437 of an arbitrary number of elliptic curve points
1438 \sum scalars[i]*points[i],
1439 optionally including the generator defined for the EC_GROUP:
1440 scalar*generator + \sum scalars[i]*points[i].
1442 EC_POINT_mul is a simple wrapper function for the typical case
1443 that the point list has just one item (besides the optional
1447 *) First EC_METHODs for curves over GF(p):
1449 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
1450 operations and provides various method functions that can also
1451 operate with faster implementations of modular arithmetic.
1453 EC_GFp_mont_method() reuses most functions that are part of
1454 EC_GFp_simple_method, but uses Montgomery arithmetic.
1456 [Bodo Moeller; point addition and point doubling
1457 implementation directly derived from source code provided by
1458 Lenka Fibikova <fibikova@exp-math.uni-essen.de>]
1460 *) Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
1461 crypto/ec/ec_lib.c):
1463 Curves are EC_GROUP objects (with an optional group generator)
1464 based on EC_METHODs that are built into the library.
1466 Points are EC_POINT objects based on EC_GROUP objects.
1468 Most of the framework would be able to handle curves over arbitrary
1469 finite fields, but as there are no obvious types for fields other
1470 than GF(p), some functions are limited to that for now.
1473 *) Add the -HTTP option to s_server. It is similar to -WWW, but requires
1474 that the file contains a complete HTTP response.
1477 *) Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
1478 change the def and num file printf format specifier from "%-40sXXX"
1479 to "%-39s XXX". The latter will always guarantee a space after the
1480 field while the former will cause them to run together if the field
1481 is 40 of more characters long.
1484 *) Constify the cipher and digest 'method' functions and structures
1485 and modify related functions to take constant EVP_MD and EVP_CIPHER
1489 *) Hide BN_CTX structure details in bn_lcl.h instead of publishing them
1490 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
1493 *) Modify EVP_Digest*() routines so they now return values. Although the
1494 internal software routines can never fail additional hardware versions
1498 *) Clean up crypto/err/err.h and change some error codes to avoid conflicts:
1500 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
1501 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
1504 ERR_R_NESTED_ASN1_ERROR
1506 ERR_R_MISSING_ASN1_EOS
1507 were 4 .. 9, conflicting with
1508 ERR_LIB_RSA (= ERR_R_RSA_LIB)
1510 ERR_LIB_PEM (= ERR_R_PEM_LIB).
1511 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
1513 Add new error code 'ERR_R_INTERNAL_ERROR'.
1516 *) Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
1520 *) New option '-subj arg' for 'openssl req' and 'openssl ca'. This
1521 sets the subject name for a new request or supersedes the
1522 subject name in a given request. Formats that can be parsed are
1523 'CN=Some Name, OU=myOU, C=IT'
1525 'CN=Some Name/OU=myOU/C=IT'.
1527 Add options '-batch' and '-verbose' to 'openssl req'.
1528 [Massimiliano Pala <madwolf@hackmasters.net>]
1530 *) Introduce the possibility to access global variables through
1531 functions on platform were that's the best way to handle exporting
1532 global variables in shared libraries. To enable this functionality,
1533 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
1534 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
1535 is normally done by Configure or something similar).
1537 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
1538 in the source file (foo.c) like this:
1540 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
1541 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
1543 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
1544 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
1546 OPENSSL_DECLARE_GLOBAL(int,foo);
1547 #define foo OPENSSL_GLOBAL_REF(foo)
1548 OPENSSL_DECLARE_GLOBAL(double,bar);
1549 #define bar OPENSSL_GLOBAL_REF(bar)
1551 The #defines are very important, and therefore so is including the
1552 header file everywhere where the defined globals are used.
1554 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
1555 of ASN.1 items, but that structure is a bit different.
1557 The largest change is in util/mkdef.pl which has been enhanced with
1558 better and easier to understand logic to choose which symbols should
1559 go into the Windows .def files as well as a number of fixes and code
1560 cleanup (among others, algorithm keywords are now sorted
1561 lexicographically to avoid constant rewrites).
1564 *) In BN_div() keep a copy of the sign of 'num' before writing the
1565 result to 'rm' because if rm==num the value will be overwritten
1566 and produce the wrong result if 'num' is negative: this caused
1567 problems with BN_mod() and BN_nnmod().
1570 *) Function OCSP_request_verify(). This checks the signature on an
1571 OCSP request and verifies the signer certificate. The signer
1572 certificate is just checked for a generic purpose and OCSP request
1576 *) Add OCSP_check_validity() function to check the validity of OCSP
1577 responses. OCSP responses are prepared in real time and may only
1578 be a few seconds old. Simply checking that the current time lies
1579 between thisUpdate and nextUpdate max reject otherwise valid responses
1580 caused by either OCSP responder or client clock inaccuracy. Instead
1581 we allow thisUpdate and nextUpdate to fall within a certain period of
1582 the current time. The age of the response can also optionally be
1583 checked. Two new options -validity_period and -status_age added to
1587 *) If signature or public key algorithm is unrecognized print out its
1588 OID rather that just UNKNOWN.
1591 *) Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
1592 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
1593 ID to be generated from the issuer certificate alone which can then be
1594 passed to OCSP_id_issuer_cmp().
1597 *) New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
1598 ASN1 modules to export functions returning ASN1_ITEM pointers
1599 instead of the ASN1_ITEM structures themselves. This adds several
1600 new macros which allow the underlying ASN1 function/structure to
1601 be accessed transparently. As a result code should not use ASN1_ITEM
1602 references directly (such as &X509_it) but instead use the relevant
1603 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
1604 use of the new ASN1 code on platforms where exporting structures
1605 is problematical (for example in shared libraries) but exporting
1606 functions returning pointers to structures is not.
1609 *) Add support for overriding the generation of SSL/TLS session IDs.
1610 These callbacks can be registered either in an SSL_CTX or per SSL.
1611 The purpose of this is to allow applications to control, if they wish,
1612 the arbitrary values chosen for use as session IDs, particularly as it
1613 can be useful for session caching in multiple-server environments. A
1614 command-line switch for testing this (and any client code that wishes
1615 to use such a feature) has been added to "s_server".
1616 [Geoff Thorpe, Lutz Jaenicke]
1618 *) Modify mkdef.pl to recognise and parse preprocessor conditionals
1619 of the form '#if defined(...) || defined(...) || ...' and
1620 '#if !defined(...) && !defined(...) && ...'. This also avoids
1621 the growing number of special cases it was previously handling.
1624 *) Make all configuration macros available for application by making
1625 sure they are available in opensslconf.h, by giving them names starting
1626 with "OPENSSL_" to avoid conflicts with other packages and by making
1627 sure e_os2.h will cover all platform-specific cases together with
1629 Additionally, it is now possible to define configuration/platform-
1630 specific names (called "system identities"). In the C code, these
1631 are prefixed with "OPENSSL_SYSNAME_". e_os2.h will create another
1632 macro with the name beginning with "OPENSSL_SYS_", which is determined
1633 from "OPENSSL_SYSNAME_*" or compiler-specific macros depending on
1637 *) New option -set_serial to 'req' and 'x509' this allows the serial
1638 number to use to be specified on the command line. Previously self
1639 signed certificates were hard coded with serial number 0 and the
1640 CA options of 'x509' had to use a serial number in a file which was
1644 *) New options to 'ca' utility to support V2 CRL entry extensions.
1645 Currently CRL reason, invalidity date and hold instruction are
1646 supported. Add new CRL extensions to V3 code and some new objects.
1649 *) New function EVP_CIPHER_CTX_set_padding() this is used to
1650 disable standard block padding (aka PKCS#5 padding) in the EVP
1651 API, which was previously mandatory. This means that the data is
1652 not padded in any way and so the total length much be a multiple
1653 of the block size, otherwise an error occurs.
1656 *) Initial (incomplete) OCSP SSL support.
1659 *) New function OCSP_parse_url(). This splits up a URL into its host,
1660 port and path components: primarily to parse OCSP URLs. New -url
1661 option to ocsp utility.
1664 *) New nonce behavior. The return value of OCSP_check_nonce() now
1665 reflects the various checks performed. Applications can decide
1666 whether to tolerate certain situations such as an absent nonce
1667 in a response when one was present in a request: the ocsp application
1668 just prints out a warning. New function OCSP_add1_basic_nonce()
1669 this is to allow responders to include a nonce in a response even if
1670 the request is nonce-less.
1673 *) Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
1674 skipped when using openssl x509 multiple times on a single input file,
1675 e.g. "(openssl x509 -out cert1; openssl x509 -out cert2) <certs".
1678 *) Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
1679 set string type: to handle setting ASN1_TIME structures. Fix ca
1680 utility to correctly initialize revocation date of CRLs.
1683 *) New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
1684 the clients preferred ciphersuites and rather use its own preferences.
1685 Should help to work around M$ SGC (Server Gated Cryptography) bug in
1686 Internet Explorer by ensuring unchanged hash method during stepup.
1687 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
1690 *) Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
1691 to aes and add a new 'exist' option to print out symbols that don't
1695 *) Additional options to ocsp utility to allow flags to be set and
1696 additional certificates supplied.
1699 *) Add the option -VAfile to 'openssl ocsp', so the user can give the
1700 OCSP client a number of certificate to only verify the response
1704 *) Update Rijndael code to version 3.0 and change EVP AES ciphers to
1705 handle the new API. Currently only ECB, CBC modes supported. Add new
1708 Add TLS AES ciphersuites as described in RFC3268, "Advanced
1709 Encryption Standard (AES) Ciphersuites for Transport Layer
1710 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
1711 not enabled by default and were not part of the "ALL" ciphersuite
1712 alias because they were not yet official; they could be
1713 explicitly requested by specifying the "AESdraft" ciphersuite
1714 group alias. In the final release of OpenSSL 0.9.7, the group
1715 alias is called "AES" and is part of "ALL".)
1716 [Ben Laurie, Steve Henson, Bodo Moeller]
1718 *) New function OCSP_copy_nonce() to copy nonce value (if present) from
1719 request to response.
1722 *) Functions for OCSP responders. OCSP_request_onereq_count(),
1723 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
1724 extract information from a certificate request. OCSP_response_create()
1725 creates a response and optionally adds a basic response structure.
1726 OCSP_basic_add1_status() adds a complete single response to a basic
1727 response and returns the OCSP_SINGLERESP structure just added (to allow
1728 extensions to be included for example). OCSP_basic_add1_cert() adds a
1729 certificate to a basic response and OCSP_basic_sign() signs a basic
1730 response with various flags. New helper functions ASN1_TIME_check()
1731 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
1732 (converts ASN1_TIME to GeneralizedTime).
1735 *) Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
1736 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
1737 structure from a certificate. X509_pubkey_digest() digests the public_key
1738 contents: this is used in various key identifiers.
1741 *) Make sk_sort() tolerate a NULL argument.
1742 [Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>]
1744 *) New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
1745 passed by the function are trusted implicitly. If any of them signed the
1746 response then it is assumed to be valid and is not verified.
1749 *) In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
1750 to data. This was previously part of the PKCS7 ASN1 code. This
1751 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
1752 [Steve Henson, reported by Kenneth R. Robinette
1753 <support@securenetterm.com>]
1755 *) Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
1756 routines: without these tracing memory leaks is very painful.
1757 Fix leaks in PKCS12 and PKCS7 routines.
1760 *) Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
1761 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
1762 effectively meant GeneralizedTime would never be used. Now it
1763 is initialised to -1 but X509_time_adj() now has to check the value
1764 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
1765 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
1766 [Steve Henson, reported by Kenneth R. Robinette
1767 <support@securenetterm.com>]
1769 *) Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
1770 result in a zero length in the ASN1_INTEGER structure which was
1771 not consistent with the structure when d2i_ASN1_INTEGER() was used
1772 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
1773 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
1774 where it did not print out a minus for negative ASN1_INTEGER.
1777 *) Add summary printout to ocsp utility. The various functions which
1778 convert status values to strings have been renamed to:
1779 OCSP_response_status_str(), OCSP_cert_status_str() and
1780 OCSP_crl_reason_str() and are no longer static. New options
1781 to verify nonce values and to disable verification. OCSP response
1782 printout format cleaned up.
1785 *) Add additional OCSP certificate checks. These are those specified
1786 in RFC2560. This consists of two separate checks: the CA of the
1787 certificate being checked must either be the OCSP signer certificate
1788 or the issuer of the OCSP signer certificate. In the latter case the
1789 OCSP signer certificate must contain the OCSP signing extended key
1790 usage. This check is performed by attempting to match the OCSP
1791 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
1792 in the OCSP_CERTID structures of the response.
1795 *) Initial OCSP certificate verification added to OCSP_basic_verify()
1796 and related routines. This uses the standard OpenSSL certificate
1797 verify routines to perform initial checks (just CA validity) and
1798 to obtain the certificate chain. Then additional checks will be
1799 performed on the chain. Currently the root CA is checked to see
1800 if it is explicitly trusted for OCSP signing. This is used to set
1801 a root CA as a global signing root: that is any certificate that
1802 chains to that CA is an acceptable OCSP signing certificate.
1805 *) New '-extfile ...' option to 'openssl ca' for reading X.509v3
1806 extensions from a separate configuration file.
1807 As when reading extensions from the main configuration file,
1808 the '-extensions ...' option may be used for specifying the
1810 [Massimiliano Pala <madwolf@comune.modena.it>]
1812 *) New OCSP utility. Allows OCSP requests to be generated or
1813 read. The request can be sent to a responder and the output
1814 parsed, outputed or printed in text form. Not complete yet:
1815 still needs to check the OCSP response validity.
1818 *) New subcommands for 'openssl ca':
1819 'openssl ca -status <serial>' prints the status of the cert with
1820 the given serial number (according to the index file).
1821 'openssl ca -updatedb' updates the expiry status of certificates
1823 [Massimiliano Pala <madwolf@comune.modena.it>]
1825 *) New '-newreq-nodes' command option to CA.pl. This is like
1826 '-newreq', but calls 'openssl req' with the '-nodes' option
1827 so that the resulting key is not encrypted.
1828 [Damien Miller <djm@mindrot.org>]
1830 *) New configuration for the GNU Hurd.
1831 [Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte]
1833 *) Initial code to implement OCSP basic response verify. This
1834 is currently incomplete. Currently just finds the signer's
1835 certificate and verifies the signature on the response.
1838 *) New SSLeay_version code SSLEAY_DIR to determine the compiled-in
1839 value of OPENSSLDIR. This is available via the new '-d' option
1840 to 'openssl version', and is also included in 'openssl version -a'.
1843 *) Allowing defining memory allocation callbacks that will be given
1844 file name and line number information in additional arguments
1845 (a const char* and an int). The basic functionality remains, as
1846 well as the original possibility to just replace malloc(),
1847 realloc() and free() by functions that do not know about these
1848 additional arguments. To register and find out the current
1849 settings for extended allocation functions, the following
1850 functions are provided:
1852 CRYPTO_set_mem_ex_functions
1853 CRYPTO_set_locked_mem_ex_functions
1854 CRYPTO_get_mem_ex_functions
1855 CRYPTO_get_locked_mem_ex_functions
1857 These work the same way as CRYPTO_set_mem_functions and friends.
1858 CRYPTO_get_[locked_]mem_functions now writes 0 where such an
1859 extended allocation function is enabled.
1860 Similarly, CRYPTO_get_[locked_]mem_ex_functions writes 0 where
1861 a conventional allocation function is enabled.
1862 [Richard Levitte, Bodo Moeller]
1864 *) Finish off removing the remaining LHASH function pointer casts.
1865 There should no longer be any prototype-casting required when using
1866 the LHASH abstraction, and any casts that remain are "bugs". See
1867 the callback types and macros at the head of lhash.h for details
1868 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
1871 *) Add automatic query of EGD sockets in RAND_poll() for the unix variant.
1872 If /dev/[u]random devices are not available or do not return enough
1873 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
1875 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
1876 /etc/entropy will be queried once each in this sequence, quering stops
1877 when enough entropy was collected without querying more sockets.
1880 *) Change the Unix RAND_poll() variant to be able to poll several
1881 random devices, as specified by DEVRANDOM, until a sufficient amount
1882 of data has been collected. We spend at most 10 ms on each file
1883 (select timeout) and read in non-blocking mode. DEVRANDOM now
1884 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
1885 (previously it was just the string "/dev/urandom"), so on typical
1886 platforms the 10 ms delay will never occur.
1887 Also separate out the Unix variant to its own file, rand_unix.c.
1888 For VMS, there's a currently-empty rand_vms.c.
1891 *) Move OCSP client related routines to ocsp_cl.c. These
1892 provide utility functions which an application needing
1893 to issue a request to an OCSP responder and analyse the
1894 response will typically need: as opposed to those which an
1895 OCSP responder itself would need which will be added later.
1897 OCSP_request_sign() signs an OCSP request with an API similar
1898 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
1899 response. OCSP_response_get1_basic() extracts basic response
1900 from response. OCSP_resp_find_status(): finds and extracts status
1901 information from an OCSP_CERTID structure (which will be created
1902 when the request structure is built). These are built from lower
1903 level functions which work on OCSP_SINGLERESP structures but
1904 wont normally be used unless the application wishes to examine
1905 extensions in the OCSP response for example.
1907 Replace nonce routines with a pair of functions.
1908 OCSP_request_add1_nonce() adds a nonce value and optionally
1909 generates a random value. OCSP_check_nonce() checks the
1910 validity of the nonce in an OCSP response.
1913 *) Change function OCSP_request_add() to OCSP_request_add0_id().
1914 This doesn't copy the supplied OCSP_CERTID and avoids the
1915 need to free up the newly created id. Change return type
1916 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
1917 This can then be used to add extensions to the request.
1918 Deleted OCSP_request_new(), since most of its functionality
1919 is now in OCSP_REQUEST_new() (and the case insensitive name
1920 clash) apart from the ability to set the request name which
1921 will be added elsewhere.
1924 *) Update OCSP API. Remove obsolete extensions argument from
1925 various functions. Extensions are now handled using the new
1926 OCSP extension code. New simple OCSP HTTP function which
1927 can be used to send requests and parse the response.
1930 *) Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
1931 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
1932 uses the special reorder version of SET OF to sort the attributes
1933 and reorder them to match the encoded order. This resolves a long
1934 standing problem: a verify on a PKCS7 structure just after signing
1935 it used to fail because the attribute order did not match the
1936 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
1937 it uses the received order. This is necessary to tolerate some broken
1938 software that does not order SET OF. This is handled by encoding
1939 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
1940 to produce the required SET OF.
1943 *) Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
1944 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
1945 files to get correct declarations of the ASN.1 item variables.
1948 *) Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
1949 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
1950 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
1951 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
1952 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
1953 ASN1_ITEM and no wrapper functions.
1956 *) New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
1957 replace the old function pointer based I/O routines. Change most of
1958 the *_d2i_bio() and *_d2i_fp() functions to use these.
1961 *) Enhance mkdef.pl to be more accepting about spacing in C preprocessor
1962 lines, recognice more "algorithms" that can be deselected, and make
1963 it complain about algorithm deselection that isn't recognised.
1966 *) New ASN1 functions to handle dup, sign, verify, digest, pack and
1967 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
1968 to use new functions. Add NO_ASN1_OLD which can be set to remove
1969 some old style ASN1 functions: this can be used to determine if old
1970 code will still work when these eventually go away.
1973 *) New extension functions for OCSP structures, these follow the
1974 same conventions as certificates and CRLs.
1977 *) New function X509V3_add1_i2d(). This automatically encodes and
1978 adds an extension. Its behaviour can be customised with various
1979 flags to append, replace or delete. Various wrappers added for
1980 certifcates and CRLs.
1983 *) Fix to avoid calling the underlying ASN1 print routine when
1984 an extension cannot be parsed. Correct a typo in the
1985 OCSP_SERVICELOC extension. Tidy up print OCSP format.
1988 *) Make mkdef.pl parse some of the ASN1 macros and add apropriate
1989 entries for variables.
1992 *) Add functionality to apps/openssl.c for detecting locking
1993 problems: As the program is single-threaded, all we have
1994 to do is register a locking callback using an array for
1995 storing which locks are currently held by the program.
1998 *) Use a lock around the call to CRYPTO_get_ex_new_index() in
1999 SSL_get_ex_data_X509_STORE_idx(), which is used in
2000 ssl_verify_cert_chain() and thus can be called at any time
2001 during TLS/SSL handshakes so that thread-safety is essential.
2002 Unfortunately, the ex_data design is not at all suited
2003 for multi-threaded use, so it probably should be abolished.
2006 *) Added Broadcom "ubsec" ENGINE to OpenSSL.
2007 [Broadcom, tweaked and integrated by Geoff Thorpe]
2009 *) Move common extension printing code to new function
2010 X509V3_print_extensions(). Reorganise OCSP print routines and
2011 implement some needed OCSP ASN1 functions. Add OCSP extensions.
2014 *) New function X509_signature_print() to remove duplication in some
2018 *) Add a special meaning when SET OF and SEQUENCE OF flags are both
2019 set (this was treated exactly the same as SET OF previously). This
2020 is used to reorder the STACK representing the structure to match the
2021 encoding. This will be used to get round a problem where a PKCS7
2022 structure which was signed could not be verified because the STACK
2023 order did not reflect the encoded order.
2026 *) Reimplement the OCSP ASN1 module using the new code.
2029 *) Update the X509V3 code to permit the use of an ASN1_ITEM structure
2030 for its ASN1 operations. The old style function pointers still exist
2031 for now but they will eventually go away.
2034 *) Merge in replacement ASN1 code from the ASN1 branch. This almost
2035 completely replaces the old ASN1 functionality with a table driven
2036 encoder and decoder which interprets an ASN1_ITEM structure describing
2037 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
2038 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
2039 has also been converted to the new form.
2042 *) Change BN_mod_exp_recp so that negative moduli are tolerated
2043 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
2044 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
2045 for negative moduli.
2048 *) Fix BN_uadd and BN_usub: Always return non-negative results instead
2049 of not touching the result's sign bit.
2052 *) BN_div bugfix: If the result is 0, the sign (res->neg) must not be
2056 *) Changed the LHASH code to use prototypes for callbacks, and created
2057 macros to declare and implement thin (optionally static) functions
2058 that provide type-safety and avoid function pointer casting for the
2059 type-specific callbacks.
2062 *) Added Kerberos Cipher Suites to be used with TLS, as written in
2064 [Veers Staats <staatsvr@asc.hpc.mil>,
2065 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte]
2067 *) Reformat the FAQ so the different questions and answers can be divided
2068 in sections depending on the subject.
2071 *) Have the zlib compression code load ZLIB.DLL dynamically under
2075 *) New function BN_mod_sqrt for computing square roots modulo a prime
2076 (using the probabilistic Tonelli-Shanks algorithm unless
2077 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
2078 be handled deterministically).
2079 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
2081 *) Make BN_mod_inverse faster by explicitly handling small quotients
2082 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
2083 512 bits], about 30% for larger ones [1024 or 2048 bits].)
2086 *) New function BN_kronecker.
2089 *) Fix BN_gcd so that it works on negative inputs; the result is
2090 positive unless both parameters are zero.
2091 Previously something reasonably close to an infinite loop was
2092 possible because numbers could be growing instead of shrinking
2093 in the implementation of Euclid's algorithm.
2096 *) Fix BN_is_word() and BN_is_one() macros to take into account the
2097 sign of the number in question.
2099 Fix BN_is_word(a,w) to work correctly for w == 0.
2101 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
2102 because its test if the absolute value of 'a' equals 'w'.
2103 Note that BN_abs_is_word does *not* handle w == 0 reliably;
2104 it exists mostly for use in the implementations of BN_is_zero(),
2105 BN_is_one(), and BN_is_word().
2108 *) New function BN_swap.
2111 *) Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
2112 the exponentiation functions are more likely to produce reasonable
2113 results on negative inputs.
2116 *) Change BN_mod_mul so that the result is always non-negative.
2117 Previously, it could be negative if one of the factors was negative;
2118 I don't think anyone really wanted that behaviour.
2121 *) Move BN_mod_... functions into new file crypto/bn/bn_mod.c
2122 (except for exponentiation, which stays in crypto/bn/bn_exp.c,
2123 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c)
2124 and add new functions:
2133 BN_mod_lshift1_quick
2137 These functions always generate non-negative results.
2139 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r
2140 such that |m| < r < 0, BN_nnmod will output rem + |m| instead).
2142 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as
2143 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b]
2144 be reduced modulo m.
2145 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
2148 The following entry accidentily appeared in the CHANGES file
2149 distributed with OpenSSL 0.9.7. The modifications described in
2150 it do *not* apply to OpenSSL 0.9.7.
2152 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
2153 was actually never needed) and in BN_mul(). The removal in BN_mul()
2154 required a small change in bn_mul_part_recursive() and the addition
2155 of the functions bn_cmp_part_words(), bn_sub_part_words() and
2156 bn_add_part_words(), which do the same thing as bn_cmp_words(),
2157 bn_sub_words() and bn_add_words() except they take arrays with
2162 *) In 'openssl passwd', verify passwords read from the terminal
2163 unless the '-salt' option is used (which usually means that
2164 verification would just waste user's time since the resulting
2165 hash is going to be compared with some given password hash)
2166 or the new '-noverify' option is used.
2168 This is an incompatible change, but it does not affect
2169 non-interactive use of 'openssl passwd' (passwords on the command
2170 line, '-stdin' option, '-in ...' option) and thus should not
2174 *) Remove all references to RSAref, since there's no more need for it.
2177 *) Make DSO load along a path given through an environment variable
2178 (SHLIB_PATH) with shl_load().
2181 *) Constify the ENGINE code as a result of BIGNUM constification.
2182 Also constify the RSA code and most things related to it. In a
2183 few places, most notable in the depth of the ASN.1 code, ugly
2184 casts back to non-const were required (to be solved at a later
2188 *) Make it so the openssl application has all engines loaded by default.
2191 *) Constify the BIGNUM routines a little more.
2194 *) Add the following functions:
2196 ENGINE_load_cswift()
2198 ENGINE_load_atalla()
2200 ENGINE_load_builtin_engines()
2202 That way, an application can itself choose if external engines that
2203 are built-in in OpenSSL shall ever be used or not. The benefit is
2204 that applications won't have to be linked with libdl or other dso
2205 libraries unless it's really needed.
2207 Changed 'openssl engine' to load all engines on demand.
2208 Changed the engine header files to avoid the duplication of some
2209 declarations (they differed!).
2212 *) 'openssl engine' can now list capabilities.
2215 *) Better error reporting in 'openssl engine'.
2218 *) Never call load_dh_param(NULL) in s_server.
2221 *) Add engine application. It can currently list engines by name and
2222 identity, and test if they are actually available.
2225 *) Improve RPM specification file by forcing symbolic linking and making
2226 sure the installed documentation is also owned by root.root.
2227 [Damien Miller <djm@mindrot.org>]
2229 *) Give the OpenSSL applications more possibilities to make use of
2230 keys (public as well as private) handled by engines.
2233 *) Add OCSP code that comes from CertCo.
2236 *) Add VMS support for the Rijndael code.
2239 *) Added untested support for Nuron crypto accelerator.
2242 *) Add support for external cryptographic devices. This code was
2243 previously distributed separately as the "engine" branch.
2244 [Geoff Thorpe, Richard Levitte]
2246 *) Rework the filename-translation in the DSO code. It is now possible to
2247 have far greater control over how a "name" is turned into a filename
2248 depending on the operating environment and any oddities about the
2249 different shared library filenames on each system.
2252 *) Support threads on FreeBSD-elf in Configure.
2255 *) Fix for SHA1 assembly problem with MASM: it produces
2256 warnings about corrupt line number information when assembling
2257 with debugging information. This is caused by the overlapping
2259 [Bernd Matthes <mainbug@celocom.de>, Steve Henson]
2262 NCONF_get_number() has no error checking at all. As a replacement,
2263 NCONF_get_number_e() is defined (_e for "error checking") and is
2264 promoted strongly. The old NCONF_get_number is kept around for
2265 binary backward compatibility.
2266 Make it possible for methods to load from something other than a BIO,
2267 by providing a function pointer that is given a name instead of a BIO.
2268 For example, this could be used to load configuration data from an
2272 *) Fix for non blocking accept BIOs. Added new I/O special reason
2273 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
2274 with non blocking I/O was not possible because no retry code was
2275 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
2279 *) Added the beginnings of Rijndael support.
2282 *) Fix for bug in DirectoryString mask setting. Add support for
2283 X509_NAME_print_ex() in 'req' and X509_print_ex() function
2284 to allow certificate printing to more controllable, additional
2285 'certopt' option to 'x509' to allow new printing options to be
2289 *) Clean old EAY MD5 hack from e_os.h.
2292 Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
2294 *) New function OPENSSL_cleanse(), which is used to cleanse a section of
2295 memory from it's contents. This is done with a counter that will
2296 place alternating values in each byte. This can be used to solve
2297 two issues: 1) the removal of calls to memset() by highly optimizing
2298 compilers, and 2) cleansing with other values than 0, since those can
2299 be read through on certain media, for example a swap space on disk.
2302 *) Bugfix: client side session caching did not work with external caching,
2303 because the session->cipher setting was not restored when reloading
2304 from the external cache. This problem was masked, when
2305 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
2306 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
2309 *) Fix client_certificate (ssl/s2_clnt.c): The permissible total
2310 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
2311 [Zeev Lieber <zeev-l@yahoo.com>]
2313 *) Undo an undocumented change introduced in 0.9.6e which caused
2314 repeated calls to OpenSSL_add_all_ciphers() and
2315 OpenSSL_add_all_digests() to be ignored, even after calling
2319 *) Change the default configuration reader to deal with last line not
2320 being properly terminated.
2323 *) Change X509_NAME_cmp() so it applies the special rules on handling
2324 DN values that are of type PrintableString, as well as RDNs of type
2325 emailAddress where the value has the type ia5String.
2326 [stefank@valicert.com via Richard Levitte]
2328 *) Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
2329 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
2330 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
2331 the bitwise-OR of the two for use by the majority of applications
2332 wanting this behaviour, and update the docs. The documented
2333 behaviour and actual behaviour were inconsistent and had been
2334 changing anyway, so this is more a bug-fix than a behavioural
2336 [Geoff Thorpe, diagnosed by Nadav Har'El]
2338 *) Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
2339 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
2342 *) Fix initialization code race conditions in
2343 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
2344 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
2345 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
2346 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
2347 ssl2_get_cipher_by_char(),
2348 ssl3_get_cipher_by_char().
2349 [Patrick McCormick <patrick@tellme.com>, Bodo Moeller]
2351 *) Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
2352 the cached sessions are flushed, as the remove_cb() might use ex_data
2353 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
2354 (see [openssl.org #212]).
2355 [Geoff Thorpe, Lutz Jaenicke]
2357 *) Fix typo in OBJ_txt2obj which incorrectly passed the content
2358 length, instead of the encoding length to d2i_ASN1_OBJECT.
2361 Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
2363 *) [In 0.9.6g-engine release:]
2364 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use '_stdcall').
2365 [Lynn Gazis <lgazis@rainbow.com>]
2367 Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
2369 *) Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
2370 and get fix the header length calculation.
2371 [Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
2372 Alon Kantor <alonk@checkpoint.com> (and others),
2375 *) Use proper error handling instead of 'assertions' in buffer
2376 overflow checks added in 0.9.6e. This prevents DoS (the
2377 assertions could call abort()).
2378 [Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller]
2380 Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
2382 *) Add various sanity checks to asn1_get_length() to reject
2383 the ASN1 length bytes if they exceed sizeof(long), will appear
2384 negative or the content length exceeds the length of the
2386 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
2388 *) Fix cipher selection routines: ciphers without encryption had no flags
2389 for the cipher strength set and where therefore not handled correctly
2390 by the selection routines (PR #130).
2393 *) Fix EVP_dsa_sha macro.
2397 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
2398 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
2399 that was added in OpenSSL 0.9.6d.
2401 As the countermeasure turned out to be incompatible with some
2402 broken SSL implementations, the new option is part of SSL_OP_ALL.
2403 SSL_OP_ALL is usually employed when compatibility with weird SSL
2404 implementations is desired (e.g. '-bugs' option to 's_client' and
2405 's_server'), so the new option is automatically set in many
2409 *) Changes in security patch:
2411 Changes marked "(CHATS)" were sponsored by the Defense Advanced
2412 Research Projects Agency (DARPA) and Air Force Research Laboratory,
2413 Air Force Materiel Command, USAF, under agreement number
2416 *) Add various sanity checks to asn1_get_length() to reject
2417 the ASN1 length bytes if they exceed sizeof(long), will appear
2418 negative or the content length exceeds the length of the
2419 supplied buffer. (CAN-2002-0659)
2420 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
2422 *) Assertions for various potential buffer overflows, not known to
2424 [Ben Laurie (CHATS)]
2426 *) Various temporary buffers to hold ASCII versions of integers were
2427 too small for 64 bit platforms. (CAN-2002-0655)
2428 [Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>
2430 *) Remote buffer overflow in SSL3 protocol - an attacker could
2431 supply an oversized session ID to a client. (CAN-2002-0656)
2432 [Ben Laurie (CHATS)]
2434 *) Remote buffer overflow in SSL2 protocol - an attacker could
2435 supply an oversized client master key. (CAN-2002-0656)
2436 [Ben Laurie (CHATS)]
2438 Changes between 0.9.6c and 0.9.6d [9 May 2002]
2440 *) Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
2441 encoded as NULL) with id-dsa-with-sha1.
2442 [Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller]
2444 *) Check various X509_...() return values in apps/req.c.
2445 [Nils Larsch <nla@trustcenter.de>]
2447 *) Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
2448 an end-of-file condition would erronously be flagged, when the CRLF
2449 was just at the end of a processed block. The bug was discovered when
2450 processing data through a buffering memory BIO handing the data to a
2451 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
2452 <ptsekov@syntrex.com> and Nedelcho Stanev.
2455 *) Implement a countermeasure against a vulnerability recently found
2456 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
2457 before application data chunks to avoid the use of known IVs
2458 with data potentially chosen by the attacker.
2461 *) Fix length checks in ssl3_get_client_hello().
2464 *) TLS/SSL library bugfix: use s->s3->in_read_app_data differently
2465 to prevent ssl3_read_internal() from incorrectly assuming that
2466 ssl3_read_bytes() found application data while handshake
2467 processing was enabled when in fact s->s3->in_read_app_data was
2468 merely automatically cleared during the initial handshake.
2469 [Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>]
2471 *) Fix object definitions for Private and Enterprise: they were not
2472 recognized in their shortname (=lowercase) representation. Extend
2473 obj_dat.pl to issue an error when using undefined keywords instead
2474 of silently ignoring the problem (Svenning Sorensen
2475 <sss@sss.dnsalias.net>).
2478 *) Fix DH_generate_parameters() so that it works for 'non-standard'
2479 generators, i.e. generators other than 2 and 5. (Previously, the
2480 code did not properly initialise the 'add' and 'rem' values to
2481 BN_generate_prime().)
2483 In the new general case, we do not insist that 'generator' is
2484 actually a primitive root: This requirement is rather pointless;
2485 a generator of the order-q subgroup is just as good, if not
2489 *) Map new X509 verification errors to alerts. Discovered and submitted by
2490 Tom Wu <tom@arcot.com>.
2493 *) Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
2494 returning non-zero before the data has been completely received
2495 when using non-blocking I/O.
2496 [Bodo Moeller; problem pointed out by John Hughes]
2498 *) Some of the ciphers missed the strength entry (SSL_LOW etc).
2499 [Ben Laurie, Lutz Jaenicke]
2501 *) Fix bug in SSL_clear(): bad sessions were not removed (found by
2502 Yoram Zahavi <YoramZ@gilian.com>).
2505 *) Add information about CygWin 1.3 and on, and preserve proper
2506 configuration for the versions before that.
2507 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
2509 *) Make removal from session cache (SSL_CTX_remove_session()) more robust:
2510 check whether we deal with a copy of a session and do not delete from
2511 the cache in this case. Problem reported by "Izhar Shoshani Levi"
2512 <izhar@checkpoint.com>.
2515 *) Do not store session data into the internal session cache, if it
2516 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
2517 flag is set). Proposed by Aslam <aslam@funk.com>.
2520 *) Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
2524 *) [In 0.9.6d-engine release:]
2525 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
2526 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
2528 *) Add the configuration target linux-s390x.
2529 [Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte]
2531 *) The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
2532 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
2533 variable as an indication that a ClientHello message has been
2534 received. As the flag value will be lost between multiple
2535 invocations of ssl3_accept when using non-blocking I/O, the
2536 function may not be aware that a handshake has actually taken
2537 place, thus preventing a new session from being added to the
2540 To avoid this problem, we now set s->new_session to 2 instead of
2541 using a local variable.
2542 [Lutz Jaenicke, Bodo Moeller]
2544 *) Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
2545 if the SSL_R_LENGTH_MISMATCH error is detected.
2546 [Geoff Thorpe, Bodo Moeller]
2548 *) New 'shared_ldflag' column in Configure platform table.
2551 *) Fix EVP_CIPHER_mode macro.
2552 ["Dan S. Camper" <dan@bti.net>]
2554 *) Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
2555 type, we must throw them away by setting rr->length to 0.
2556 [D P Chang <dpc@qualys.com>]
2558 Changes between 0.9.6b and 0.9.6c [21 dec 2001]
2560 *) Fix BN_rand_range bug pointed out by Dominikus Scherkl
2561 <Dominikus.Scherkl@biodata.com>. (The previous implementation
2562 worked incorrectly for those cases where range = 10..._2 and
2563 3*range is two bits longer than range.)
2566 *) Only add signing time to PKCS7 structures if it is not already
2570 *) Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
2571 OBJ_ld_ce should be OBJ_id_ce.
2572 Also some ip-pda OIDs in crypto/objects/objects.txt were
2573 incorrect (cf. RFC 3039).
2574 [Matt Cooper, Frederic Giudicelli, Bodo Moeller]
2576 *) Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
2577 returns early because it has nothing to do.
2578 [Andy Schneider <andy.schneider@bjss.co.uk>]
2580 *) [In 0.9.6c-engine release:]
2581 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
2582 [Andy Schneider <andy.schneider@bjss.co.uk>]
2584 *) [In 0.9.6c-engine release:]
2585 Add support for Cryptographic Appliance's keyserver technology.
2586 (Use engine 'keyclient')
2587 [Cryptographic Appliances and Geoff Thorpe]
2589 *) Add a configuration entry for OS/390 Unix. The C compiler 'c89'
2590 is called via tools/c89.sh because arguments have to be
2591 rearranged (all '-L' options must appear before the first object
2593 [Richard Shapiro <rshapiro@abinitio.com>]
2595 *) [In 0.9.6c-engine release:]
2596 Add support for Broadcom crypto accelerator cards, backported
2598 [Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox]
2600 *) [In 0.9.6c-engine release:]
2601 Add support for SureWare crypto accelerator cards from
2602 Baltimore Technologies. (Use engine 'sureware')
2603 [Baltimore Technologies and Mark Cox]
2605 *) [In 0.9.6c-engine release:]
2606 Add support for crypto accelerator cards from Accelerated
2607 Encryption Processing, www.aep.ie. (Use engine 'aep')
2608 [AEP Inc. and Mark Cox]
2610 *) Add a configuration entry for gcc on UnixWare.
2611 [Gary Benson <gbenson@redhat.com>]
2613 *) Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
2614 messages are stored in a single piece (fixed-length part and
2615 variable-length part combined) and fix various bugs found on the way.
2618 *) Disable caching in BIO_gethostbyname(), directly use gethostbyname()
2619 instead. BIO_gethostbyname() does not know what timeouts are
2620 appropriate, so entries would stay in cache even when they have
2622 [Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>
2624 *) Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
2625 faced with a pathologically small ClientHello fragment that does
2626 not contain client_version: Instead of aborting with an error,
2627 simply choose the highest available protocol version (i.e.,
2628 TLS 1.0 unless it is disabled). In practice, ClientHello
2629 messages are never sent like this, but this change gives us
2630 strictly correct behaviour at least for TLS.
2633 *) Fix SSL handshake functions and SSL_clear() such that SSL_clear()
2634 never resets s->method to s->ctx->method when called from within
2635 one of the SSL handshake functions.
2636 [Bodo Moeller; problem pointed out by Niko Baric]
2638 *) In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
2639 (sent using the client's version number) if client_version is
2640 smaller than the protocol version in use. Also change
2641 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
2642 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
2643 the client will at least see that alert.
2646 *) Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
2650 *) Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
2651 client receives HelloRequest while in a handshake.
2652 [Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>]
2654 *) Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
2655 should end in 'break', not 'goto end' which circuments various
2656 cleanups done in state SSL_ST_OK. But session related stuff
2657 must be disabled for SSL_ST_OK in the case that we just sent a
2660 Also avoid some overhead by not calling ssl_init_wbio_buffer()
2661 before just sending a HelloRequest.
2662 [Bodo Moeller, Eric Rescorla <ekr@rtfm.com>]
2664 *) Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
2665 reveal whether illegal block cipher padding was found or a MAC
2666 verification error occured. (Neither SSLerr() codes nor alerts
2667 are directly visible to potential attackers, but the information
2668 may leak via logfiles.)
2670 Similar changes are not required for the SSL 2.0 implementation
2671 because the number of padding bytes is sent in clear for SSL 2.0,
2672 and the extra bytes are just ignored. However ssl/s2_pkt.c
2673 failed to verify that the purported number of padding bytes is in
2677 *) Add OpenUNIX-8 support including shared libraries
2678 (Boyd Lynn Gerber <gerberb@zenez.com>).
2681 *) Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
2682 'wristwatch attack' using huge encoding parameters (cf.
2683 James H. Manger's CRYPTO 2001 paper). Note that the
2684 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
2685 encoding parameters and hence was not vulnerable.
2688 *) BN_sqr() bug fix.
2689 [Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>]
2691 *) Rabin-Miller test analyses assume uniformly distributed witnesses,
2692 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
2693 followed by modular reduction.
2694 [Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>]
2696 *) Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
2697 equivalent based on BN_pseudo_rand() instead of BN_rand().
2700 *) s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
2701 This function was broken, as the check for a new client hello message
2702 to handle SGC did not allow these large messages.
2703 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
2706 *) Add alert descriptions for TLSv1 to SSL_alert_desc_string[_long]().
2709 *) Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
2710 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
2713 *) Rework the configuration and shared library support for Tru64 Unix.
2714 The configuration part makes use of modern compiler features and
2715 still retains old compiler behavior for those that run older versions
2716 of the OS. The shared library support part includes a variant that
2717 uses the RPATH feature, and is available through the special
2718 configuration target "alpha-cc-rpath", which will never be selected
2720 [Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte]
2722 *) In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
2723 with the same message size as in ssl3_get_certificate_request().
2724 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
2725 messages might inadvertently be reject as too long.
2726 [Petr Lampa <lampa@fee.vutbr.cz>]
2728 *) Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
2731 *) Modified SSL library such that the verify_callback that has been set
2732 specificly for an SSL object with SSL_set_verify() is actually being
2733 used. Before the change, a verify_callback set with this function was
2734 ignored and the verify_callback() set in the SSL_CTX at the time of
2735 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
2736 to allow the necessary settings.
2739 *) Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
2740 explicitly to NULL, as at least on Solaris 8 this seems not always to be
2741 done automatically (in contradiction to the requirements of the C
2742 standard). This made problems when used from OpenSSH.
2745 *) In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
2746 dh->length and always used
2748 BN_rand_range(priv_key, dh->p).
2750 BN_rand_range() is not necessary for Diffie-Hellman, and this
2751 specific range makes Diffie-Hellman unnecessarily inefficient if
2752 dh->length (recommended exponent length) is much smaller than the
2753 length of dh->p. We could use BN_rand_range() if the order of
2754 the subgroup was stored in the DH structure, but we only have
2759 BN_rand(priv_key, l, ...)
2761 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
2767 RSA_eay_public_encrypt
2768 RSA_eay_private_decrypt
2769 RSA_eay_private_encrypt (signing)
2770 RSA_eay_public_decrypt (signature verification)
2772 (default implementations for RSA_public_encrypt,
2773 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
2774 always reject numbers >= n.
2777 *) In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
2778 to synchronize access to 'locking_thread'. This is necessary on
2779 systems where access to 'locking_thread' (an 'unsigned long'
2780 variable) is not atomic.
2783 *) In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
2784 *before* setting the 'crypto_lock_rand' flag. The previous code had
2785 a race condition if 0 is a valid thread ID.
2786 [Travis Vitek <vitek@roguewave.com>]
2788 *) Add support for shared libraries under Irix.
2789 [Albert Chin-A-Young <china@thewrittenword.com>]
2791 *) Add configuration option to build on Linux on both big-endian and
2793 [Ralf Baechle <ralf@uni-koblenz.de>]
2795 *) Add the possibility to create shared libraries on HP-UX.
2798 Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
2800 *) Change ssleay_rand_bytes (crypto/rand/md_rand.c)
2801 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
2802 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
2803 PRNG state recovery was possible based on the output of
2804 one PRNG request appropriately sized to gain knowledge on
2805 'md' followed by enough consecutive 1-byte PRNG requests
2806 to traverse all of 'state'.
2808 1. When updating 'md_local' (the current thread's copy of 'md')
2809 during PRNG output generation, hash all of the previous
2810 'md_local' value, not just the half used for PRNG output.
2812 2. Make the number of bytes from 'state' included into the hash
2813 independent from the number of PRNG bytes requested.
2815 The first measure alone would be sufficient to avoid
2816 Markku-Juhani's attack. (Actually it had never occurred
2817 to me that the half of 'md_local' used for chaining was the
2818 half from which PRNG output bytes were taken -- I had always
2819 assumed that the secret half would be used.) The second
2820 measure makes sure that additional data from 'state' is never
2821 mixed into 'md_local' in small portions; this heuristically
2822 further strengthens the PRNG.
2825 *) Fix crypto/bn/asm/mips3.s.
2828 *) When only the key is given to "enc", the IV is undefined. Print out
2829 an error message in this case.
2832 *) Handle special case when X509_NAME is empty in X509 printing routines.
2835 *) In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
2836 positive and less than q.
2839 *) Don't change *pointer in CRYPTO_add_lock() is add_lock_callback is
2840 used: it isn't thread safe and the add_lock_callback should handle
2842 [Paul Rose <Paul.Rose@bridge.com>]
2844 *) Verify that incoming data obeys the block size in
2845 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
2849 [Ulf Möller, Bodo Möller]
2851 *) The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
2852 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
2853 when fixing the server behaviour for backwards-compatible 'client
2854 hello' messages. (Note that the attack is impractical against
2855 SSL 3.0 and TLS 1.0 anyway because length and version checking
2856 means that the probability of guessing a valid ciphertext is
2857 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
2860 Before 0.9.5, the countermeasure (hide the error by generating a
2861 random 'decryption result') did not work properly because
2862 ERR_clear_error() was missing, meaning that SSL_get_error() would
2863 detect the supposedly ignored error.
2865 Both problems are now fixed.
2868 *) In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
2869 (previously it was 1024).
2872 *) Fix for compatibility mode trust settings: ignore trust settings
2873 unless some valid trust or reject settings are present.
2876 *) Fix for blowfish EVP: its a variable length cipher.
2879 *) Fix various bugs related to DSA S/MIME verification. Handle missing
2880 parameters in DSA public key structures and return an error in the
2881 DSA routines if parameters are absent.
2884 *) In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
2885 in the current directory if neither $RANDFILE nor $HOME was set.
2886 RAND_file_name() in 0.9.6a returned NULL in this case. This has
2887 caused some confusion to Windows users who haven't defined $HOME.
2888 Thus RAND_file_name() is changed again: e_os.h can define a
2889 DEFAULT_HOME, which will be used if $HOME is not set.
2890 For Windows, we use "C:"; on other platforms, we still require
2891 environment variables.
2893 *) Move 'if (!initialized) RAND_poll()' into regions protected by
2894 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
2895 having multiple threads call RAND_poll() concurrently.
2898 *) In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
2899 combination of a flag and a thread ID variable.
2900 Otherwise while one thread is in ssleay_rand_bytes (which sets the
2901 flag), *other* threads can enter ssleay_add_bytes without obeying
2902 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
2903 that they do not hold after the first thread unsets add_do_not_lock).
2906 *) Change bctest again: '-x' expressions are not available in all
2910 Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
2912 *) Fix a couple of memory leaks in PKCS7_dataDecode()
2913 [Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>]
2915 *) Change Configure and Makefiles to provide EXE_EXT, which will contain
2916 the default extension for executables, if any. Also, make the perl
2917 scripts that use symlink() to test if it really exists and use "cp"
2918 if it doesn't. All this made OpenSSL compilable and installable in
2922 *) Fix for asn1_GetSequence() for indefinite length constructed data.
2923 If SEQUENCE is length is indefinite just set c->slen to the total
2924 amount of data available.
2925 [Steve Henson, reported by shige@FreeBSD.org]
2926 [This change does not apply to 0.9.7.]
2928 *) Change bctest to avoid here-documents inside command substitution
2929 (workaround for FreeBSD /bin/sh bug).
2930 For compatibility with Ultrix, avoid shell functions (introduced
2931 in the bctest version that searches along $PATH).
2934 *) Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
2935 with des_encrypt() defined on some operating systems, like Solaris
2939 *) Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
2940 On the Importance of Eliminating Errors in Cryptographic
2941 Computations, J. Cryptology 14 (2001) 2, 101-119,
2942 http://theory.stanford.edu/~dabo/papers/faults.ps.gz).
2945 *) MIPS assembler BIGNUM division bug fix.
2948 *) Disabled incorrect Alpha assembler code.
2951 *) Fix PKCS#7 decode routines so they correctly update the length
2952 after reading an EOC for the EXPLICIT tag.
2954 [This change does not apply to 0.9.7.]
2956 *) Fix bug in PKCS#12 key generation routines. This was triggered
2957 if a 3DES key was generated with a 0 initial byte. Include
2958 PKCS12_BROKEN_KEYGEN compilation option to retain the old
2959 (but broken) behaviour.
2962 *) Enhance bctest to search for a working bc along $PATH and print
2964 [Tim Rice <tim@multitalents.net> via Richard Levitte]
2966 *) Fix memory leaks in err.c: free err_data string if necessary;
2967 don't write to the wrong index in ERR_set_error_data.
2970 *) Implement ssl23_peek (analogous to ssl23_read), which previously
2974 *) Replace rdtsc with _emit statements for VC++ version 5.
2975 [Jeremy Cooper <jeremy@baymoo.org>]
2977 *) Make it possible to reuse SSLv2 sessions.
2980 *) In copy_email() check for >= 0 as a return value for
2981 X509_NAME_get_index_by_NID() since 0 is a valid index.
2982 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>]
2984 *) Avoid coredump with unsupported or invalid public keys by checking if
2985 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
2986 PKCS7_verify() fails with non detached data.
2989 *) Don't use getenv in library functions when run as setuid/setgid.
2990 New function OPENSSL_issetugid().
2993 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
2994 due to incorrect handling of multi-threading:
2996 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
2998 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
3000 3. Count how many times MemCheck_off() has been called so that
3001 nested use can be treated correctly. This also avoids
3002 inband-signalling in the previous code (which relied on the
3003 assumption that thread ID 0 is impossible).
3006 *) Add "-rand" option also to s_client and s_server.
3009 *) Fix CPU detection on Irix 6.x.
3010 [Kurt Hockenbury <khockenb@stevens-tech.edu> and
3011 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
3013 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME
3016 [This change does not apply to 0.9.7.]
3018 *) Use the cached encoding of an X509_NAME structure rather than
3019 copying it. This is apparently the reason for the libsafe "errors"
3020 but the code is actually correct.
3023 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
3024 Bleichenbacher's DSA attack.
3025 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
3026 to be set and top=0 forces the highest bit to be set; top=-1 is new
3027 and leaves the highest bit random.
3028 [Ulf Moeller, Bodo Moeller]
3030 *) In the NCONF_...-based implementations for CONF_... queries
3031 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
3032 a temporary CONF structure with the data component set to NULL
3033 (which gives segmentation faults in lh_retrieve).
3034 Instead, use NULL for the CONF pointer in CONF_get_string and
3035 CONF_get_number (which may use environment variables) and directly
3036 return NULL from CONF_get_section.
3039 *) Fix potential buffer overrun for EBCDIC.
3042 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign
3043 keyUsage if basicConstraints absent for a CA.
3046 *) Make SMIME_write_PKCS7() write mail header values with a format that
3047 is more generally accepted (no spaces before the semicolon), since
3048 some programs can't parse those values properly otherwise. Also make
3049 sure BIO's that break lines after each write do not create invalid
3053 *) Make the CRL encoding routines work with empty SEQUENCE OF. The
3054 macros previously used would not encode an empty SEQUENCE OF
3055 and break the signature.
3057 [This change does not apply to 0.9.7.]
3059 *) Zero the premaster secret after deriving the master secret in
3063 *) Add some EVP_add_digest_alias registrations (as found in
3064 OpenSSL_add_all_digests()) to SSL_library_init()
3065 aka OpenSSL_add_ssl_algorithms(). This provides improved
3066 compatibility with peers using X.509 certificates
3067 with unconventional AlgorithmIdentifier OIDs.
3070 *) Fix for Irix with NO_ASM.
3071 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
3073 *) ./config script fixes.
3074 [Ulf Moeller, Richard Levitte]
3076 *) Fix 'openssl passwd -1'.
3079 *) Change PKCS12_key_gen_asc() so it can cope with non null
3080 terminated strings whose length is passed in the passlen
3081 parameter, for example from PEM callbacks. This was done
3082 by adding an extra length parameter to asc2uni().
3083 [Steve Henson, reported by <oddissey@samsung.co.kr>]
3085 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
3086 call failed, free the DSA structure.
3089 *) Fix to uni2asc() to cope with zero length Unicode strings.
3090 These are present in some PKCS#12 files.
3093 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
3094 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
3095 when writing a 32767 byte record.
3096 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>]
3098 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c),
3099 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}.
3101 (RSA objects have a reference count access to which is protected
3102 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
3103 so they are meant to be shared between threads.)
3104 [Bodo Moeller, Geoff Thorpe; original patch submitted by
3105 "Reddie, Steven" <Steven.Reddie@ca.com>]
3107 *) Fix a deadlock in CRYPTO_mem_leaks().
3110 *) Use better test patterns in bntest.
3113 *) rand_win.c fix for Borland C.
3116 *) BN_rshift bugfix for n == 0.
3119 *) Add a 'bctest' script that checks for some known 'bc' bugs
3120 so that 'make test' does not abort just because 'bc' is broken.
3123 *) Store verify_result within SSL_SESSION also for client side to
3124 avoid potential security hole. (Re-used sessions on the client side
3125 always resulted in verify_result==X509_V_OK, not using the original
3126 result of the server certificate verification.)
3129 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type
3130 SSL3_RT_APPLICATION_DATA, return 0.
3131 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
3135 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
3136 releases, have been re-implemented by renaming the previous
3137 implementations of ssl2_read and ssl3_read to ssl2_read_internal
3138 and ssl3_read_internal, respectively, and adding 'peek' parameters
3139 to them. The new ssl[23]_{read,peek} functions are calls to
3140 ssl[23]_read_internal with the 'peek' flag set appropriately.
3141 A 'peek' parameter has also been added to ssl3_read_bytes, which
3142 does the actual work for ssl3_read_internal.
3145 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
3146 the method-specific "init()" handler. Also clean up ex_data after
3147 calling the method-specific "finish()" handler. Previously, this was
3148 happening the other way round.
3151 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
3152 The previous value, 12, was not always sufficient for BN_mod_exp().
3155 *) Make sure that shared libraries get the internal name engine with
3156 the full version number and not just 0. This should mark the
3157 shared libraries as not backward compatible. Of course, this should
3158 be changed again when we can guarantee backward binary compatibility.
3161 *) Fix typo in get_cert_by_subject() in by_dir.c
3162 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>]
3164 *) Rework the system to generate shared libraries:
3166 - Make note of the expected extension for the shared libraries and
3167 if there is a need for symbolic links from for example libcrypto.so.0
3168 to libcrypto.so.0.9.7. There is extended info in Configure for
3171 - Make as few rebuilds of the shared libraries as possible.
3173 - Still avoid linking the OpenSSL programs with the shared libraries.
3175 - When installing, install the shared libraries separately from the
3179 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument.
3181 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
3182 and not in SSL_clear because the latter is also used by the
3183 accept/connect functions; previously, the settings made by
3184 SSL_set_read_ahead would be lost during the handshake.
3185 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>]
3187 *) Correct util/mkdef.pl to be selective about disabled algorithms.
3188 Previously, it would create entries for disableed algorithms no
3192 *) Added several new manual pages for SSL_* function.
3195 Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
3197 *) In ssl23_get_client_hello, generate an error message when faced
3198 with an initial SSL 3.0/TLS record that is too small to contain the
3199 first two bytes of the ClientHello message, i.e. client_version.
3200 (Note that this is a pathologic case that probably has never happened
3201 in real life.) The previous approach was to use the version number
3202 from the record header as a substitute; but our protocol choice
3203 should not depend on that one because it is not authenticated
3204 by the Finished messages.
3207 *) More robust randomness gathering functions for Windows.
3208 [Jeffrey Altman <jaltman@columbia.edu>]
3210 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
3211 not set then we don't setup the error code for issuer check errors
3212 to avoid possibly overwriting other errors which the callback does
3213 handle. If an application does set the flag then we assume it knows
3214 what it is doing and can handle the new informational codes
3218 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
3219 a general "ANY" type, as such it should be able to decode anything
3220 including tagged types. However it didn't check the class so it would
3221 wrongly interpret tagged types in the same way as their universal
3222 counterpart and unknown types were just rejected. Changed so that the
3223 tagged and unknown types are handled in the same way as a SEQUENCE:
3224 that is the encoding is stored intact. There is also a new type
3225 "V_ASN1_OTHER" which is used when the class is not universal, in this
3226 case we have no idea what the actual type is so we just lump them all
3230 *) On VMS, stdout may very well lead to a file that is written to
3231 in a record-oriented fashion. That means that every write() will
3232 write a separate record, which will be read separately by the
3233 programs trying to read from it. This can be very confusing.
3235 The solution is to put a BIO filter in the way that will buffer
3236 text until a linefeed is reached, and then write everything a
3237 line at a time, so every record written will be an actual line,
3238 not chunks of lines and not (usually doesn't happen, but I've
3239 seen it once) several lines in one record. BIO_f_linebuffer() is
3242 Currently, it's a VMS-only method, because that's where it has
3243 been tested well enough.
3246 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
3247 it can return incorrect results.
3248 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
3249 but it was in 0.9.6-beta[12].)
3252 *) Disable the check for content being present when verifying detached
3253 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
3254 include zero length content when signing messages.
3257 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
3258 BIO_ctrl (for BIO pairs).
3261 *) Add DSO method for VMS.
3264 *) Bug fix: Montgomery multiplication could produce results with the
3268 *) Add RPM specification openssl.spec and modify it to build three
3269 packages. The default package contains applications, application
3270 documentation and run-time libraries. The devel package contains
3271 include files, static libraries and function documentation. The
3272 doc package contains the contents of the doc directory. The original
3273 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
3276 *) Add a large number of documentation files for many SSL routines.
3277 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
3279 *) Add a configuration entry for Sony News 4.
3280 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>]
3282 *) Don't set the two most significant bits to one when generating a
3283 random number < q in the DSA library.
3286 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
3287 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
3288 the underlying transport is blocking) if a handshake took place.
3289 (The default behaviour is needed by applications such as s_client
3290 and s_server that use select() to determine when to use SSL_read;
3291 but for applications that know in advance when to expect data, it
3292 just makes things more complicated.)
3295 *) Add RAND_egd_bytes(), which gives control over the number of bytes read
3299 *) Add a few more EBCDIC conditionals that make `req' and `x509'
3300 work better on such systems.
3301 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
3303 *) Add two demo programs for PKCS12_parse() and PKCS12_create().
3304 Update PKCS12_parse() so it copies the friendlyName and the
3305 keyid to the certificates aux info.
3308 *) Fix bug in PKCS7_verify() which caused an infinite loop
3309 if there was more than one signature.
3310 [Sven Uszpelkat <su@celocom.de>]
3312 *) Major change in util/mkdef.pl to include extra information
3313 about each symbol, as well as presentig variables as well
3314 as functions. This change means that there's n more need
3315 to rebuild the .num files when some algorithms are excluded.
3318 *) Allow the verify time to be set by an application,
3319 rather than always using the current time.
3322 *) Phase 2 verify code reorganisation. The certificate
3323 verify code now looks up an issuer certificate by a
3324 number of criteria: subject name, authority key id
3325 and key usage. It also verifies self signed certificates
3326 by the same criteria. The main comparison function is
3327 X509_check_issued() which performs these checks.
3329 Lot of changes were necessary in order to support this
3330 without completely rewriting the lookup code.
3332 Authority and subject key identifier are now cached.
3334 The LHASH 'certs' is X509_STORE has now been replaced
3335 by a STACK_OF(X509_OBJECT). This is mainly because an
3336 LHASH can't store or retrieve multiple objects with
3337 the same hash value.
3339 As a result various functions (which were all internal
3340 use only) have changed to handle the new X509_STORE
3341 structure. This will break anything that messed round
3342 with X509_STORE internally.
3344 The functions X509_STORE_add_cert() now checks for an
3345 exact match, rather than just subject name.
3347 The X509_STORE API doesn't directly support the retrieval
3348 of multiple certificates matching a given criteria, however
3349 this can be worked round by performing a lookup first
3350 (which will fill the cache with candidate certificates)
3351 and then examining the cache for matches. This is probably
3352 the best we can do without throwing out X509_LOOKUP
3353 entirely (maybe later...).
3355 The X509_VERIFY_CTX structure has been enhanced considerably.
3357 All certificate lookup operations now go via a get_issuer()
3358 callback. Although this currently uses an X509_STORE it
3359 can be replaced by custom lookups. This is a simple way
3360 to bypass the X509_STORE hackery necessary to make this
3361 work and makes it possible to use more efficient techniques
3362 in future. A very simple version which uses a simple
3363 STACK for its trusted certificate store is also provided
3364 using X509_STORE_CTX_trusted_stack().
3366 The verify_cb() and verify() callbacks now have equivalents
3367 in the X509_STORE_CTX structure.
3369 X509_STORE_CTX also has a 'flags' field which can be used
3370 to customise the verify behaviour.
3373 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
3374 excludes S/MIME capabilities.
3377 *) When a certificate request is read in keep a copy of the
3378 original encoding of the signed data and use it when outputing
3379 again. Signatures then use the original encoding rather than
3380 a decoded, encoded version which may cause problems if the
3381 request is improperly encoded.
3384 *) For consistency with other BIO_puts implementations, call
3385 buffer_write(b, ...) directly in buffer_puts instead of calling
3388 In BIO_puts, increment b->num_write as in BIO_write.
3389 [Peter.Sylvester@EdelWeb.fr]
3391 *) Fix BN_mul_word for the case where the word is 0. (We have to use
3392 BN_zero, we may not return a BIGNUM with an array consisting of
3396 *) Avoid calling abort() from within the library when problems are
3397 detected, except if preprocessor symbols have been defined
3398 (such as REF_CHECK, BN_DEBUG etc.).
3401 *) New openssl application 'rsautl'. This utility can be
3402 used for low level RSA operations. DER public key
3403 BIO/fp routines also added.
3406 *) New Configure entry and patches for compiling on QNX 4.
3407 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
3409 *) A demo state-machine implementation was sponsored by
3410 Nuron (http://www.nuron.com/) and is now available in
3411 demos/state_machine.
3414 *) New options added to the 'dgst' utility for signature
3415 generation and verification.
3418 *) Unrecognized PKCS#7 content types are now handled via a
3419 catch all ASN1_TYPE structure. This allows unsupported
3420 types to be stored as a "blob" and an application can
3421 encode and decode it manually.
3424 *) Fix various signed/unsigned issues to make a_strex.c
3426 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
3428 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
3429 length if passed a buffer. ASN1_INTEGER_to_BN failed
3430 if passed a NULL BN and its argument was negative.
3431 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
3433 *) Modification to PKCS#7 encoding routines to output definite
3434 length encoding. Since currently the whole structures are in
3435 memory there's not real point in using indefinite length
3436 constructed encoding. However if OpenSSL is compiled with
3437 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
3440 *) Added BIO_vprintf() and BIO_vsnprintf().
3443 *) Added more prefixes to parse for in the the strings written
3444 through a logging bio, to cover all the levels that are available
3445 through syslog. The prefixes are now:
3447 PANIC, EMERG, EMR => LOG_EMERG
3448 ALERT, ALR => LOG_ALERT
3449 CRIT, CRI => LOG_CRIT
3450 ERROR, ERR => LOG_ERR
3451 WARNING, WARN, WAR => LOG_WARNING
3452 NOTICE, NOTE, NOT => LOG_NOTICE
3453 INFO, INF => LOG_INFO
3454 DEBUG, DBG => LOG_DEBUG
3456 and as before, if none of those prefixes are present at the
3457 beginning of the string, LOG_ERR is chosen.
3459 On Win32, the LOG_* levels are mapped according to this:
3461 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
3462 LOG_WARNING => EVENTLOG_WARNING_TYPE
3463 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
3467 *) Made it possible to reconfigure with just the configuration
3468 argument "reconf" or "reconfigure". The command line arguments
3469 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
3470 and are retrieved from there when reconfiguring.
3474 [Assar Westerlund <assar@sics.se>, Richard Levitte]
3476 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
3479 *) The obj_dat.pl script was messing up the sorting of object
3480 names. The reason was that it compared the quoted version
3481 of strings as a result "OCSP" > "OCSP Signing" because
3482 " > SPACE. Changed script to store unquoted versions of
3483 names and add quotes on output. It was also omitting some
3484 names from the lookup table if they were given a default
3485 value (that is if SN is missing it is given the same
3486 value as LN and vice versa), these are now added on the
3487 grounds that if an object has a name we should be able to
3488 look it up. Finally added warning output when duplicate
3489 short or long names are found.
3492 *) Changes needed for Tandem NSK.
3493 [Scott Uroff <scott@xypro.com>]
3495 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
3496 RSA_padding_check_SSLv23(), special padding was never detected
3497 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
3498 version rollback attacks was not effective.
3500 In s23_clnt.c, don't use special rollback-attack detection padding
3501 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
3502 client; similarly, in s23_srvr.c, don't do the rollback check if
3503 SSL 2.0 is the only protocol enabled in the server.
3506 *) Make it possible to get hexdumps of unprintable data with 'openssl
3507 asn1parse'. By implication, the functions ASN1_parse_dump() and
3508 BIO_dump_indent() are added.
3511 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
3512 these print out strings and name structures based on various
3513 flags including RFC2253 support and proper handling of
3514 multibyte characters. Added options to the 'x509' utility
3515 to allow the various flags to be set.
3518 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
3519 Also change the functions X509_cmp_current_time() and
3520 X509_gmtime_adj() work with an ASN1_TIME structure,
3521 this will enable certificates using GeneralizedTime in validity
3522 dates to be checked.
3525 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
3526 negative public key encodings) on by default,
3527 NO_NEG_PUBKEY_BUG can be set to disable it.
3530 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
3531 content octets. An i2c_ASN1_OBJECT is unnecessary because
3532 the encoding can be trivially obtained from the structure.
3535 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
3536 not read locks (CRYPTO_r_[un]lock).
3539 *) A first attempt at creating official support for shared
3540 libraries through configuration. I've kept it so the
3541 default is static libraries only, and the OpenSSL programs
3542 are always statically linked for now, but there are
3543 preparations for dynamic linking in place.
3544 This has been tested on Linux and Tru64.
3547 *) Randomness polling function for Win9x, as described in:
3548 Peter Gutmann, Software Generation of Practically Strong
3552 *) Fix so PRNG is seeded in req if using an already existing
3556 *) New options to smime application. -inform and -outform
3557 allow alternative formats for the S/MIME message including
3558 PEM and DER. The -content option allows the content to be
3559 specified separately. This should allow things like Netscape
3560 form signing output easier to verify.
3563 *) Fix the ASN1 encoding of tags using the 'long form'.
3566 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
3567 STRING types. These convert content octets to and from the
3568 underlying type. The actual tag and length octets are
3569 already assumed to have been read in and checked. These
3570 are needed because all other string types have virtually
3571 identical handling apart from the tag. By having versions
3572 of the ASN1 functions that just operate on content octets
3573 IMPLICIT tagging can be handled properly. It also allows
3574 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
3575 and ASN1_INTEGER are identical apart from the tag.
3578 *) Change the handling of OID objects as follows:
3580 - New object identifiers are inserted in objects.txt, following
3581 the syntax given in objects.README.
3582 - objects.pl is used to process obj_mac.num and create a new
3584 - obj_dat.pl is used to create a new obj_dat.h, using the data in
3587 This is currently kind of a hack, and the perl code in objects.pl
3588 isn't very elegant, but it works as I intended. The simplest way
3589 to check that it worked correctly is to look in obj_dat.h and
3590 check the array nid_objs and make sure the objects haven't moved
3591 around (this is important!). Additions are OK, as well as
3592 consistent name changes.
3595 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
3598 *) Addition of the command line parameter '-rand file' to 'openssl req'.
3599 The given file adds to whatever has already been seeded into the
3600 random pool through the RANDFILE configuration file option or
3601 environment variable, or the default random state file.
3604 *) mkstack.pl now sorts each macro group into lexical order.
3605 Previously the output order depended on the order the files
3606 appeared in the directory, resulting in needless rewriting
3610 *) Patches to make OpenSSL compile under Win32 again. Mostly
3611 work arounds for the VC++ problem that it treats func() as
3612 func(void). Also stripped out the parts of mkdef.pl that
3613 added extra typesafe functions: these no longer exist.
3616 *) Reorganisation of the stack code. The macros are now all
3617 collected in safestack.h . Each macro is defined in terms of
3618 a "stack macro" of the form SKM_<name>(type, a, b). The
3619 DEBUG_SAFESTACK is now handled in terms of function casts,
3620 this has the advantage of retaining type safety without the
3621 use of additional functions. If DEBUG_SAFESTACK is not defined
3622 then the non typesafe macros are used instead. Also modified the
3623 mkstack.pl script to handle the new form. Needs testing to see
3624 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
3625 the default if no major problems. Similar behaviour for ASN1_SET_OF
3626 and PKCS12_STACK_OF.
3629 *) When some versions of IIS use the 'NET' form of private key the
3630 key derivation algorithm is different. Normally MD5(password) is
3631 used as a 128 bit RC4 key. In the modified case
3632 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
3633 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
3634 as the old Netscape_RSA functions except they have an additional
3635 'sgckey' parameter which uses the modified algorithm. Also added
3636 an -sgckey command line option to the rsa utility. Thanks to
3637 Adrian Peck <bertie@ncipher.com> for posting details of the modified
3638 algorithm to openssl-dev.
3641 *) The evp_local.h macros were using 'c.##kname' which resulted in
3642 invalid expansion on some systems (SCO 5.0.5 for example).
3643 Corrected to 'c.kname'.
3644 [Phillip Porch <root@theporch.com>]
3646 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
3647 a STACK of email addresses from a certificate or request, these look
3648 in the subject name and the subject alternative name extensions and
3649 omit any duplicate addresses.
3652 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
3653 This makes DSA verification about 2 % faster.
3656 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
3657 (meaning that now 2^5 values will be precomputed, which is only 4 KB
3658 plus overhead for 1024 bit moduli).
3659 This makes exponentiations about 0.5 % faster for 1024 bit
3660 exponents (as measured by "openssl speed rsa2048").
3663 *) Rename memory handling macros to avoid conflicts with other
3665 Malloc => OPENSSL_malloc
3666 Malloc_locked => OPENSSL_malloc_locked
3667 Realloc => OPENSSL_realloc
3668 Free => OPENSSL_free
3671 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
3672 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
3675 *) CygWin32 support.
3676 [John Jarvie <jjarvie@newsguy.com>]
3678 *) The type-safe stack code has been rejigged. It is now only compiled
3679 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
3680 by default all type-specific stack functions are "#define"d back to
3681 standard stack functions. This results in more streamlined output
3682 but retains the type-safety checking possibilities of the original
3686 *) The STACK code has been cleaned up, and certain type declarations
3687 that didn't make a lot of sense have been brought in line. This has
3688 also involved a cleanup of sorts in safestack.h to more correctly
3689 map type-safe stack functions onto their plain stack counterparts.
3690 This work has also resulted in a variety of "const"ifications of
3691 lots of the code, especially "_cmp" operations which should normally
3692 be prototyped with "const" parameters anyway.
3695 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
3696 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
3697 (The PRNG state consists of two parts, the large pool 'state' and 'md',
3698 where all of 'md' is used each time the PRNG is used, but 'state'
3699 is used only indexed by a cyclic counter. As entropy may not be
3700 well distributed from the beginning, 'md' is important as a
3701 chaining variable. However, the output function chains only half
3702 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
3703 all of 'md', and seeding with STATE_SIZE dummy bytes will result
3704 in all of 'state' being rewritten, with the new values depending
3705 on virtually all of 'md'. This overcomes the 80 bit limitation.)
3708 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
3709 the handshake is continued after ssl_verify_cert_chain();
3710 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
3711 can lead to 'unexplainable' connection aborts later.
3712 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
3714 *) Major EVP API cipher revision.
3715 Add hooks for extra EVP features. This allows various cipher
3716 parameters to be set in the EVP interface. Support added for variable
3717 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
3718 setting of RC2 and RC5 parameters.
3720 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
3723 Remove lots of duplicated code from the EVP library. For example *every*
3724 cipher init() function handles the 'iv' in the same way according to the
3725 cipher mode. They also all do nothing if the 'key' parameter is NULL and
3726 for CFB and OFB modes they zero ctx->num.
3728 New functionality allows removal of S/MIME code RC2 hack.
3730 Most of the routines have the same form and so can be declared in terms
3733 By shifting this to the top level EVP_CipherInit() it can be removed from
3734 all individual ciphers. If the cipher wants to handle IVs or keys
3735 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
3738 Change lots of functions like EVP_EncryptUpdate() to now return a
3739 value: although software versions of the algorithms cannot fail
3740 any installed hardware versions can.
3743 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
3744 this option is set, tolerate broken clients that send the negotiated
3745 protocol version number instead of the requested protocol version
3749 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
3750 i.e. non-zero for export ciphersuites, zero otherwise.
3751 Previous versions had this flag inverted, inconsistent with
3752 rsa_tmp_cb (..._TMP_RSA_CB).
3753 [Bodo Moeller; problem reported by Amit Chopra]
3755 *) Add missing DSA library text string. Work around for some IIS
3756 key files with invalid SEQUENCE encoding.
3759 *) Add a document (doc/standards.txt) that list all kinds of standards
3760 and so on that are implemented in OpenSSL.
3763 *) Enhance c_rehash script. Old version would mishandle certificates
3764 with the same subject name hash and wouldn't handle CRLs at all.
3765 Added -fingerprint option to crl utility, to support new c_rehash
3769 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
3772 *) Fix for SSL server purpose checking. Server checking was
3773 rejecting certificates which had extended key usage present
3774 but no ssl client purpose.
3775 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
3777 *) Make PKCS#12 code work with no password. The PKCS#12 spec
3778 is a little unclear about how a blank password is handled.
3779 Since the password in encoded as a BMPString with terminating
3780 double NULL a zero length password would end up as just the
3781 double NULL. However no password at all is different and is
3782 handled differently in the PKCS#12 key generation code. NS
3783 treats a blank password as zero length. MSIE treats it as no
3784 password on export: but it will try both on import. We now do
3785 the same: PKCS12_parse() tries zero length and no password if
3786 the password is set to "" or NULL (NULL is now a valid password:
3787 it wasn't before) as does the pkcs12 application.
3790 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
3791 perror when PEM_read_bio_X509_REQ fails, the error message must
3792 be obtained from the error queue.
3795 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
3796 it in ERR_remove_state if appropriate, and change ERR_get_state
3797 accordingly to avoid race conditions (this is necessary because
3798 thread_hash is no longer constant once set).
3801 *) Bugfix for linux-elf makefile.one.
3804 *) RSA_get_default_method() will now cause a default
3805 RSA_METHOD to be chosen if one doesn't exist already.
3806 Previously this was only set during a call to RSA_new()
3807 or RSA_new_method(NULL) meaning it was possible for
3808 RSA_get_default_method() to return NULL.
3811 *) Added native name translation to the existing DSO code
3812 that will convert (if the flag to do so is set) filenames
3813 that are sufficiently small and have no path information
3814 into a canonical native form. Eg. "blah" converted to
3815 "libblah.so" or "blah.dll" etc.
3818 *) New function ERR_error_string_n(e, buf, len) which is like
3819 ERR_error_string(e, buf), but writes at most 'len' bytes
3820 including the 0 terminator. For ERR_error_string_n, 'buf'
3822 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
3824 *) CONF library reworked to become more general. A new CONF
3825 configuration file reader "class" is implemented as well as a
3826 new functions (NCONF_*, for "New CONF") to handle it. The now
3827 old CONF_* functions are still there, but are reimplemented to
3828 work in terms of the new functions. Also, a set of functions
3829 to handle the internal storage of the configuration data is
3830 provided to make it easier to write new configuration file
3831 reader "classes" (I can definitely see something reading a
3832 configuration file in XML format, for example), called _CONF_*,
3833 or "the configuration storage API"...
3835 The new configuration file reading functions are:
3837 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
3838 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
3840 NCONF_default, NCONF_WIN32
3842 NCONF_dump_fp, NCONF_dump_bio
3844 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
3845 NCONF_new creates a new CONF object. This works in the same way
3846 as other interfaces in OpenSSL, like the BIO interface.
3847 NCONF_dump_* dump the internal storage of the configuration file,
3848 which is useful for debugging. All other functions take the same
3849 arguments as the old CONF_* functions wth the exception of the
3850 first that must be a `CONF *' instead of a `LHASH *'.
3852 To make it easer to use the new classes with the old CONF_* functions,
3853 the function CONF_set_default_method is provided.
3856 *) Add '-tls1' option to 'openssl ciphers', which was already
3857 mentioned in the documentation but had not been implemented.
3858 (This option is not yet really useful because even the additional
3859 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
3862 *) Initial DSO code added into libcrypto for letting OpenSSL (and
3863 OpenSSL-based applications) load shared libraries and bind to
3864 them in a portable way.
3865 [Geoff Thorpe, with contributions from Richard Levitte]
3867 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
3869 *) Make sure _lrotl and _lrotr are only used with MSVC.
3871 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
3872 (the default implementation of RAND_status).
3874 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
3875 to '-clrext' (= clear extensions), as intended and documented.
3876 [Bodo Moeller; inconsistency pointed out by Michael Attili
3879 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
3880 was larger than the MD block size.
3881 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
3883 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
3884 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
3885 using the passed key: if the passed key was a private key the result
3886 of X509_print(), for example, would be to print out all the private key
3890 *) des_quad_cksum() byte order bug fix.
3891 [Ulf Möller, using the problem description in krb4-0.9.7, where
3892 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
3894 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
3896 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
3898 *) For easily testing in shell scripts whether some command
3899 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
3900 returns with exit code 0 iff no command of the given name is available.
3901 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
3902 the output goes to stdout and nothing is printed to stderr.
3903 Additional arguments are always ignored.
3905 Since for each cipher there is a command of the same name,
3906 the 'no-cipher' compilation switches can be tested this way.
3908 ('openssl no-XXX' is not able to detect pseudo-commands such
3909 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
3912 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
3915 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
3916 is set; it will be thrown away anyway because each handshake creates
3918 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
3919 to parameters -- in previous versions (since OpenSSL 0.9.3) the
3920 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
3921 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
3924 *) New s_client option -ign_eof: EOF at stdin is ignored, and
3925 'Q' and 'R' lose their special meanings (quit/renegotiate).
3926 This is part of what -quiet does; unlike -quiet, -ign_eof
3927 does not suppress any output.
3930 *) Add compatibility options to the purpose and trust code. The
3931 purpose X509_PURPOSE_ANY is "any purpose" which automatically
3932 accepts a certificate or CA, this was the previous behaviour,
3933 with all the associated security issues.
3935 X509_TRUST_COMPAT is the old trust behaviour: only and
3936 automatically trust self signed roots in certificate store. A
3937 new trust setting X509_TRUST_DEFAULT is used to specify that
3938 a purpose has no associated trust setting and it should instead
3939 use the value in the default purpose.
3942 *) Fix the PKCS#8 DSA private key code so it decodes keys again
3943 and fix a memory leak.
3946 *) In util/mkerr.pl (which implements 'make errors'), preserve
3947 reason strings from the previous version of the .c file, as
3948 the default to have only downcase letters (and digits) in
3949 automatically generated reasons codes is not always appropriate.
3952 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
3953 using strerror. Previously, ERR_reason_error_string() returned
3954 library names as reason strings for SYSerr; but SYSerr is a special
3955 case where small numbers are errno values, not library numbers.
3958 *) Add '-dsaparam' option to 'openssl dhparam' application. This
3959 converts DSA parameters into DH parameters. (When creating parameters,
3960 DSA_generate_parameters is used.)
3963 *) Include 'length' (recommended exponent length) in C code generated
3964 by 'openssl dhparam -C'.
3967 *) The second argument to set_label in perlasm was already being used
3968 so couldn't be used as a "file scope" flag. Moved to third argument
3972 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
3973 instead of RAND_bytes for encryption IVs and salts.
3976 *) Include RAND_status() into RAND_METHOD instead of implementing
3977 it only for md_rand.c Otherwise replacing the PRNG by calling
3978 RAND_set_rand_method would be impossible.
3981 *) Don't let DSA_generate_key() enter an infinite loop if the random
3982 number generation fails.
3985 *) New 'rand' application for creating pseudo-random output.
3988 *) Added configuration support for Linux/IA64
3989 [Rolf Haberrecker <rolf@suse.de>]
3991 *) Assembler module support for Mingw32.
3994 *) Shared library support for HPUX (in shlib/).
3995 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
3997 *) Shared library support for Solaris gcc.
3998 [Lutz Behnke <behnke@trustcenter.de>]
4000 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
4002 *) PKCS7_encrypt() was adding text MIME headers twice because they
4003 were added manually and by SMIME_crlf_copy().
4006 *) In bntest.c don't call BN_rand with zero bits argument.
4007 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
4009 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
4010 case was implemented. This caused BN_div_recp() to fail occasionally.
4013 *) Add an optional second argument to the set_label() in the perl
4014 assembly language builder. If this argument exists and is set
4015 to 1 it signals that the assembler should use a symbol whose
4016 scope is the entire file, not just the current function. This
4017 is needed with MASM which uses the format label:: for this scope.
4018 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
4020 *) Change the ASN1 types so they are typedefs by default. Before
4021 almost all types were #define'd to ASN1_STRING which was causing
4022 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
4026 *) Change names of new functions to the new get1/get0 naming
4027 convention: After 'get1', the caller owns a reference count
4028 and has to call ..._free; 'get0' returns a pointer to some
4029 data structure without incrementing reference counters.
4030 (Some of the existing 'get' functions increment a reference
4031 counter, some don't.)
4032 Similarly, 'set1' and 'add1' functions increase reference
4033 counters or duplicate objects.
4036 *) Allow for the possibility of temp RSA key generation failure:
4037 the code used to assume it always worked and crashed on failure.
4040 *) Fix potential buffer overrun problem in BIO_printf().
4041 [Ulf Möller, using public domain code by Patrick Powell; problem
4042 pointed out by David Sacerdote <das33@cornell.edu>]
4044 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
4045 RAND_egd() and RAND_status(). In the command line application,
4046 the EGD socket can be specified like a seed file using RANDFILE
4050 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
4051 Some CAs (e.g. Verisign) distribute certificates in this form.
4054 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
4055 list to exclude them. This means that no special compilation option
4056 is needed to use anonymous DH: it just needs to be included in the
4060 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
4061 EVP_MD_type. The old functionality is available in a new macro called
4062 EVP_MD_md(). Change code that uses it and update docs.
4065 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
4066 where the 'void *' argument is replaced by a function pointer argument.
4067 Previously 'void *' was abused to point to functions, which works on
4068 many platforms, but is not correct. As these functions are usually
4069 called by macros defined in OpenSSL header files, most source code
4070 should work without changes.
4073 *) <openssl/opensslconf.h> (which is created by Configure) now contains
4074 sections with information on -D... compiler switches used for
4075 compiling the library so that applications can see them. To enable
4076 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
4077 must be defined. E.g.,
4078 #define OPENSSL_ALGORITHM_DEFINES
4079 #include <openssl/opensslconf.h>
4080 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
4081 [Richard Levitte, Ulf and Bodo Möller]
4083 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
4087 *) Change the 'other' type in certificate aux info to a STACK_OF
4088 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
4089 the required ASN1 format: arbitrary types determined by an OID.
4092 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
4093 argument to 'req'. This is not because the function is newer or
4094 better than others it just uses the work 'NEW' in the certificate
4095 request header lines. Some software needs this.
4098 *) Reorganise password command line arguments: now passwords can be
4099 obtained from various sources. Delete the PEM_cb function and make
4100 it the default behaviour: i.e. if the callback is NULL and the
4101 usrdata argument is not NULL interpret it as a null terminated pass
4102 phrase. If usrdata and the callback are NULL then the pass phrase
4103 is prompted for as usual.
4106 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
4107 the support is automatically enabled. The resulting binaries will
4108 autodetect the card and use it if present.
4109 [Ben Laurie and Compaq Inc.]
4111 *) Work around for Netscape hang bug. This sends certificate request
4112 and server done in one record. Since this is perfectly legal in the
4113 SSL/TLS protocol it isn't a "bug" option and is on by default. See
4114 the bugs/SSLv3 entry for more info.
4117 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
4120 *) Add -rand argument to smime and pkcs12 applications and read/write
4124 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
4127 *) Add command line password options to the remaining applications.
4130 *) Bug fix for BN_div_recp() for numerators with an even number of
4134 *) More tests in bntest.c, and changed test_bn output.
4137 *) ./config recognizes MacOS X now.
4140 *) Bug fix for BN_div() when the first words of num and divsor are
4141 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
4144 *) Add support for various broken PKCS#8 formats, and command line
4145 options to produce them.
4148 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
4149 get temporary BIGNUMs from a BN_CTX.
4152 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
4156 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
4157 include a #define from the old name to the new. The original intent
4158 was that statically linked binaries could for example just call
4159 SSLeay_add_all_ciphers() to just add ciphers to the table and not
4160 link with digests. This never worked becayse SSLeay_add_all_digests()
4161 and SSLeay_add_all_ciphers() were in the same source file so calling
4162 one would link with the other. They are now in separate source files.
4165 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
4168 *) Use a less unusual form of the Miller-Rabin primality test (it used
4169 a binary algorithm for exponentiation integrated into the Miller-Rabin
4170 loop, our standard modexp algorithms are faster).
4173 *) Support for the EBCDIC character set completed.
4174 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
4176 *) Source code cleanups: use const where appropriate, eliminate casts,
4177 use void * instead of char * in lhash.
4180 *) Bugfix: ssl3_send_server_key_exchange was not restartable
4181 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
4182 this the server could overwrite ephemeral keys that the client
4186 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
4187 using 50 iterations of the Rabin-Miller test.
4189 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
4190 iterations of the Rabin-Miller test as required by the appendix
4191 to FIPS PUB 186[-1]) instead of DSA_is_prime.
4192 As BN_is_prime_fasttest includes trial division, DSA parameter
4193 generation becomes much faster.
4195 This implies a change for the callback functions in DSA_is_prime
4196 and DSA_generate_parameters: The callback function is called once
4197 for each positive witness in the Rabin-Miller test, not just
4198 occasionally in the inner loop; and the parameters to the
4199 callback function now provide an iteration count for the outer
4200 loop rather than for the current invocation of the inner loop.
4201 DSA_generate_parameters additionally can call the callback
4202 function with an 'iteration count' of -1, meaning that a
4203 candidate has passed the trial division test (when q is generated
4204 from an application-provided seed, trial division is skipped).
4207 *) New function BN_is_prime_fasttest that optionally does trial
4208 division before starting the Rabin-Miller test and has
4209 an additional BN_CTX * argument (whereas BN_is_prime always
4210 has to allocate at least one BN_CTX).
4211 'callback(1, -1, cb_arg)' is called when a number has passed the
4212 trial division stage.
4215 *) Fix for bug in CRL encoding. The validity dates weren't being handled
4219 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
4222 *) New function BN_pseudo_rand().
4225 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
4226 bignum version of BN_from_montgomery() with the working code from
4227 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
4231 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
4232 made it impossible to use the same SSL_SESSION data structure in
4233 SSL2 clients in multiple threads.
4236 *) The return value of RAND_load_file() no longer counts bytes obtained
4237 by stat(). RAND_load_file(..., -1) is new and uses the complete file
4238 to seed the PRNG (previously an explicit byte count was required).
4239 [Ulf Möller, Bodo Möller]
4241 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
4242 used (char *) instead of (void *) and had casts all over the place.
4245 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
4248 *) Retain source code compatibility for BN_prime_checks macro:
4249 BN_is_prime(..., BN_prime_checks, ...) now uses
4250 BN_prime_checks_for_size to determine the appropriate number of
4251 Rabin-Miller iterations.
4254 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
4255 DH_CHECK_P_NOT_SAFE_PRIME.
4256 (Check if this is true? OpenPGP calls them "strong".)
4259 *) Merge the functionality of "dh" and "gendh" programs into a new program
4260 "dhparam". The old programs are retained for now but will handle DH keys
4261 (instead of parameters) in future.
4264 *) Make the ciphers, s_server and s_client programs check the return values
4265 when a new cipher list is set.
4268 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
4269 ciphers. Before when the 56bit ciphers were enabled the sorting was
4272 The syntax for the cipher sorting has been extended to support sorting by
4273 cipher-strength (using the strength_bits hard coded in the tables).
4274 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
4276 Fix a bug in the cipher-command parser: when supplying a cipher command
4277 string with an "undefined" symbol (neither command nor alphanumeric
4278 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
4279 an error is flagged.
4281 Due to the strength-sorting extension, the code of the
4282 ssl_create_cipher_list() function was completely rearranged. I hope that
4283 the readability was also increased :-)
4284 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
4286 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
4287 for the first serial number and places 2 in the serial number file. This
4288 avoids problems when the root CA is created with serial number zero and
4289 the first user certificate has the same issuer name and serial number
4293 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
4294 the new code. Add documentation for this stuff.
4297 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
4298 X509_*() to X509at_*() on the grounds that they don't handle X509
4299 structures and behave in an analagous way to the X509v3 functions:
4300 they shouldn't be called directly but wrapper functions should be used
4303 So we also now have some wrapper functions that call the X509at functions
4304 when passed certificate requests. (TO DO: similar things can be done with
4305 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
4306 things. Some of these need some d2i or i2d and print functionality
4307 because they handle more complex structures.)
4310 *) Add missing #ifndefs that caused missing symbols when building libssl
4311 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
4312 NO_RSA in ssl/s2*.c.
4313 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
4315 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
4316 has a return value which indicates the quality of the random data
4317 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
4318 error queue. New function RAND_pseudo_bytes() generates output that is
4319 guaranteed to be unique but not unpredictable. RAND_add is like
4320 RAND_seed, but takes an extra argument for an entropy estimate
4321 (RAND_seed always assumes full entropy).
4324 *) Do more iterations of Rabin-Miller probable prime test (specifically,
4325 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
4326 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
4327 in crypto/bn/bn_prime.c for the complete table). This guarantees a
4328 false-positive rate of at most 2^-80 for random input.
4331 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
4334 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
4335 in the 0.9.5 release), this returns the chain
4336 from an X509_CTX structure with a dup of the stack and all
4337 the X509 reference counts upped: so the stack will exist
4338 after X509_CTX_cleanup() has been called. Modify pkcs12.c
4341 Also make SSL_SESSION_print() print out the verify return
4345 *) Add manpage for the pkcs12 command. Also change the default
4346 behaviour so MAC iteration counts are used unless the new
4347 -nomaciter option is used. This improves file security and
4348 only older versions of MSIE (4.0 for example) need it.
4351 *) Honor the no-xxx Configure options when creating .DEF files.
4354 *) Add PKCS#10 attributes to field table: challengePassword,
4355 unstructuredName and unstructuredAddress. These are taken from
4356 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
4357 international characters are used.
4359 More changes to X509_ATTRIBUTE code: allow the setting of types
4360 based on strings. Remove the 'loc' parameter when adding
4361 attributes because these will be a SET OF encoding which is sorted
4365 *) Initial changes to the 'req' utility to allow request generation
4366 automation. This will allow an application to just generate a template
4367 file containing all the field values and have req construct the
4370 Initial support for X509_ATTRIBUTE handling. Stacks of these are
4371 used all over the place including certificate requests and PKCS#7
4372 structures. They are currently handled manually where necessary with
4373 some primitive wrappers for PKCS#7. The new functions behave in a
4374 manner analogous to the X509 extension functions: they allow
4375 attributes to be looked up by NID and added.
4377 Later something similar to the X509V3 code would be desirable to
4378 automatically handle the encoding, decoding and printing of the
4379 more complex types. The string types like challengePassword can
4380 be handled by the string table functions.
4382 Also modified the multi byte string table handling. Now there is
4383 a 'global mask' which masks out certain types. The table itself
4384 can use the flag STABLE_NO_MASK to ignore the mask setting: this
4385 is useful when for example there is only one permissible type
4386 (as in countryName) and using the mask might result in no valid
4390 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
4391 SSL_get_peer_finished to allow applications to obtain the latest
4392 Finished messages sent to the peer or expected from the peer,
4393 respectively. (SSL_get_peer_finished is usually the Finished message
4394 actually received from the peer, otherwise the protocol will be aborted.)
4396 As the Finished message are message digests of the complete handshake
4397 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
4398 be used for external authentication procedures when the authentication
4399 provided by SSL/TLS is not desired or is not enough.
4402 *) Enhanced support for Alpha Linux is added. Now ./config checks if
4403 the host supports BWX extension and if Compaq C is present on the
4404 $PATH. Just exploiting of the BWX extension results in 20-30%
4405 performance kick for some algorithms, e.g. DES and RC4 to mention
4406 a couple. Compaq C in turn generates ~20% faster code for MD5 and
4410 *) Add support for MS "fast SGC". This is arguably a violation of the
4411 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
4412 weak crypto and after checking the certificate is SGC a second one
4413 with strong crypto. MS SGC stops the first handshake after receiving
4414 the server certificate message and sends a second client hello. Since
4415 a server will typically do all the time consuming operations before
4416 expecting any further messages from the client (server key exchange
4417 is the most expensive) there is little difference between the two.
4419 To get OpenSSL to support MS SGC we have to permit a second client
4420 hello message after we have sent server done. In addition we have to
4421 reset the MAC if we do get this second client hello.
4424 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
4425 if a DER encoded private key is RSA or DSA traditional format. Changed
4426 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
4427 format DER encoded private key. Newer code should use PKCS#8 format which
4428 has the key type encoded in the ASN1 structure. Added DER private key
4429 support to pkcs8 application.
4432 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
4433 ciphersuites has been selected (as required by the SSL 3/TLS 1
4434 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
4435 is set, we interpret this as a request to violate the specification
4436 (the worst that can happen is a handshake failure, and 'correct'
4437 behaviour would result in a handshake failure anyway).
4440 *) In SSL_CTX_add_session, take into account that there might be multiple
4441 SSL_SESSION structures with the same session ID (e.g. when two threads
4442 concurrently obtain them from an external cache).
4443 The internal cache can handle only one SSL_SESSION with a given ID,
4444 so if there's a conflict, we now throw out the old one to achieve
4448 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
4449 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
4450 some routines that use cipher OIDs: some ciphers do not have OIDs
4451 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
4455 *) Simplify the trust setting structure and code. Now we just have
4456 two sequences of OIDs for trusted and rejected settings. These will
4457 typically have values the same as the extended key usage extension
4458 and any application specific purposes.
4460 The trust checking code now has a default behaviour: it will just
4461 check for an object with the same NID as the passed id. Functions can
4462 be provided to override either the default behaviour or the behaviour
4463 for a given id. SSL client, server and email already have functions
4464 in place for compatibility: they check the NID and also return "trusted"
4465 if the certificate is self signed.
4468 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
4469 traditional format into an EVP_PKEY structure.
4472 *) Add a password callback function PEM_cb() which either prompts for
4473 a password if usr_data is NULL or otherwise assumes it is a null
4474 terminated password. Allow passwords to be passed on command line
4475 environment or config files in a few more utilities.
4478 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
4479 keys. Add some short names for PKCS#8 PBE algorithms and allow them
4480 to be specified on the command line for the pkcs8 and pkcs12 utilities.
4481 Update documentation.
4484 *) Support for ASN1 "NULL" type. This could be handled before by using
4485 ASN1_TYPE but there wasn't any function that would try to read a NULL
4486 and produce an error if it couldn't. For compatibility we also have
4487 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
4488 don't allocate anything because they don't need to.
4491 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
4493 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
4495 *) Rebuild of the memory allocation routines used by OpenSSL code and
4496 possibly others as well. The purpose is to make an interface that
4497 provide hooks so anyone can build a separate set of allocation and
4498 deallocation routines to be used by OpenSSL, for example memory
4499 pool implementations, or something else, which was previously hard
4500 since Malloc(), Realloc() and Free() were defined as macros having
4501 the values malloc, realloc and free, respectively (except for Win32
4502 compilations). The same is provided for memory debugging code.
4503 OpenSSL already comes with functionality to find memory leaks, but
4504 this gives people a chance to debug other memory problems.
4506 With these changes, a new set of functions and macros have appeared:
4508 CRYPTO_set_mem_debug_functions() [F]
4509 CRYPTO_get_mem_debug_functions() [F]
4510 CRYPTO_dbg_set_options() [F]
4511 CRYPTO_dbg_get_options() [F]
4512 CRYPTO_malloc_debug_init() [M]
4514 The memory debug functions are NULL by default, unless the library
4515 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
4516 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
4517 gives the standard debugging functions that come with OpenSSL) or
4518 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
4519 provided by the library user) must be used. When the standard
4520 debugging functions are used, CRYPTO_dbg_set_options can be used to
4521 request additional information:
4522 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
4523 the CRYPTO_MDEBUG_xxx macro when compiling the library.
4525 Also, things like CRYPTO_set_mem_functions will always give the
4526 expected result (the new set of functions is used for allocation
4527 and deallocation) at all times, regardless of platform and compiler
4530 To finish it up, some functions that were never use in any other
4531 way than through macros have a new API and new semantic:
4534 CRYPTO_dbg_realloc()
4537 All macros of value have retained their old syntax.
4538 [Richard Levitte and Bodo Moeller]
4540 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
4541 ordering of SMIMECapabilities wasn't in "strength order" and there
4542 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
4546 *) Some ASN1 types with illegal zero length encoding (INTEGER,
4547 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
4548 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
4550 *) Merge in my S/MIME library for OpenSSL. This provides a simple
4551 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
4552 functionality to handle multipart/signed properly) and a utility
4553 called 'smime' to call all this stuff. This is based on code I
4554 originally wrote for Celo who have kindly allowed it to be
4555 included in OpenSSL.
4558 *) Add variants des_set_key_checked and des_set_key_unchecked of
4559 des_set_key (aka des_key_sched). Global variable des_check_key
4560 decides which of these is called by des_set_key; this way
4561 des_check_key behaves as it always did, but applications and
4562 the library itself, which was buggy for des_check_key == 1,
4563 have a cleaner way to pick the version they need.
4566 *) New function PKCS12_newpass() which changes the password of a
4570 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
4571 dynamic mix. In both cases the ids can be used as an index into the
4572 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
4573 functions so they accept a list of the field values and the
4574 application doesn't need to directly manipulate the X509_TRUST
4578 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
4582 *) Modify the way the V3 extension code looks up extensions. This now
4583 works in a similar way to the object code: we have some "standard"
4584 extensions in a static table which is searched with OBJ_bsearch()
4585 and the application can add dynamic ones if needed. The file
4586 crypto/x509v3/ext_dat.h now has the info: this file needs to be
4587 updated whenever a new extension is added to the core code and kept
4588 in ext_nid order. There is a simple program 'tabtest.c' which checks
4589 this. New extensions are not added too often so this file can readily
4590 be maintained manually.
4592 There are two big advantages in doing things this way. The extensions
4593 can be looked up immediately and no longer need to be "added" using
4594 X509V3_add_standard_extensions(): this function now does nothing.
4595 [Side note: I get *lots* of email saying the extension code doesn't
4596 work because people forget to call this function]
4597 Also no dynamic allocation is done unless new extensions are added:
4598 so if we don't add custom extensions there is no need to call
4599 X509V3_EXT_cleanup().
4602 *) Modify enc utility's salting as follows: make salting the default. Add a
4603 magic header, so unsalted files fail gracefully instead of just decrypting
4604 to garbage. This is because not salting is a big security hole, so people
4605 should be discouraged from doing it.
4608 *) Fixes and enhancements to the 'x509' utility. It allowed a message
4609 digest to be passed on the command line but it only used this
4610 parameter when signing a certificate. Modified so all relevant
4611 operations are affected by the digest parameter including the
4612 -fingerprint and -x509toreq options. Also -x509toreq choked if a
4613 DSA key was used because it didn't fix the digest.
4616 *) Initial certificate chain verify code. Currently tests the untrusted
4617 certificates for consistency with the verify purpose (which is set
4618 when the X509_STORE_CTX structure is set up) and checks the pathlength.
4620 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
4621 this is because it will reject chains with invalid extensions whereas
4622 every previous version of OpenSSL and SSLeay made no checks at all.
4624 Trust code: checks the root CA for the relevant trust settings. Trust
4625 settings have an initial value consistent with the verify purpose: e.g.
4626 if the verify purpose is for SSL client use it expects the CA to be
4627 trusted for SSL client use. However the default value can be changed to
4628 permit custom trust settings: one example of this would be to only trust
4629 certificates from a specific "secure" set of CAs.
4631 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
4632 which should be used for version portability: especially since the
4633 verify structure is likely to change more often now.
4635 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
4636 to set them. If not set then assume SSL clients will verify SSL servers
4639 Two new options to the verify program: -untrusted allows a set of
4640 untrusted certificates to be passed in and -purpose which sets the
4641 intended purpose of the certificate. If a purpose is set then the
4642 new chain verify code is used to check extension consistency.
4645 *) Support for the authority information access extension.
4648 *) Modify RSA and DSA PEM read routines to transparently handle
4649 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
4650 public keys in a format compatible with certificate
4651 SubjectPublicKeyInfo structures. Unfortunately there were already
4652 functions called *_PublicKey_* which used various odd formats so
4653 these are retained for compatibility: however the DSA variants were
4654 never in a public release so they have been deleted. Changed dsa/rsa
4655 utilities to handle the new format: note no releases ever handled public
4656 keys so we should be OK.
4658 The primary motivation for this change is to avoid the same fiasco
4659 that dogs private keys: there are several incompatible private key
4660 formats some of which are standard and some OpenSSL specific and
4661 require various evil hacks to allow partial transparent handling and
4662 even then it doesn't work with DER formats. Given the option anything
4663 other than PKCS#8 should be dumped: but the other formats have to
4664 stay in the name of compatibility.
4666 With public keys and the benefit of hindsight one standard format
4667 is used which works with EVP_PKEY, RSA or DSA structures: though
4668 it clearly returns an error if you try to read the wrong kind of key.
4670 Added a -pubkey option to the 'x509' utility to output the public key.
4671 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
4672 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
4673 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
4674 that do the same as the EVP_PKEY_assign_*() except they up the
4675 reference count of the added key (they don't "swallow" the
4679 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
4680 CRLs would fail if the file contained no certificates or no CRLs:
4681 added a new function to read in both types and return the number
4682 read: this means that if none are read it will be an error. The
4683 DER versions of the certificate and CRL reader would always fail
4684 because it isn't possible to mix certificates and CRLs in DER format
4685 without choking one or the other routine. Changed this to just read
4686 a certificate: this is the best we can do. Also modified the code
4687 in apps/verify.c to take notice of return codes: it was previously
4688 attempting to read in certificates from NULL pointers and ignoring
4689 any errors: this is one reason why the cert and CRL reader seemed
4690 to work. It doesn't check return codes from the default certificate
4691 routines: these may well fail if the certificates aren't installed.
4694 *) Code to support otherName option in GeneralName.
4697 *) First update to verify code. Change the verify utility
4698 so it warns if it is passed a self signed certificate:
4699 for consistency with the normal behaviour. X509_verify
4700 has been modified to it will now verify a self signed
4701 certificate if *exactly* the same certificate appears
4702 in the store: it was previously impossible to trust a
4703 single self signed certificate. This means that:
4704 openssl verify ss.pem
4705 now gives a warning about a self signed certificate but
4706 openssl verify -CAfile ss.pem ss.pem
4710 *) For servers, store verify_result in SSL_SESSION data structure
4711 (and add it to external session representation).
4712 This is needed when client certificate verifications fails,
4713 but an application-provided verification callback (set by
4714 SSL_CTX_set_cert_verify_callback) allows accepting the session
4715 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
4716 but returns 1): When the session is reused, we have to set
4717 ssl->verify_result to the appropriate error code to avoid
4719 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
4721 *) Fix a bug in the new PKCS#7 code: it didn't consider the
4722 case in PKCS7_dataInit() where the signed PKCS7 structure
4723 didn't contain any existing data because it was being created.
4724 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
4726 *) Add a salt to the key derivation routines in enc.c. This
4727 forms the first 8 bytes of the encrypted file. Also add a
4728 -S option to allow a salt to be input on the command line.
4731 *) New function X509_cmp(). Oddly enough there wasn't a function
4732 to compare two certificates. We do this by working out the SHA1
4733 hash and comparing that. X509_cmp() will be needed by the trust
4737 *) SSL_get1_session() is like SSL_get_session(), but increments
4738 the reference count in the SSL_SESSION returned.
4739 [Geoff Thorpe <geoff@eu.c2.net>]
4741 *) Fix for 'req': it was adding a null to request attributes.
4742 Also change the X509_LOOKUP and X509_INFO code to handle
4743 certificate auxiliary information.
4746 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
4750 *) Add the possibility to add extra information to the memory leak
4751 detecting output, to form tracebacks, showing from where each
4752 allocation was originated: CRYPTO_push_info("constant string") adds
4753 the string plus current file name and line number to a per-thread
4754 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
4755 is like calling CYRPTO_pop_info() until the stack is empty.
4756 Also updated memory leak detection code to be multi-thread-safe.
4759 *) Add options -text and -noout to pkcs7 utility and delete the
4760 encryption options which never did anything. Update docs.
4763 *) Add options to some of the utilities to allow the pass phrase
4764 to be included on either the command line (not recommended on
4765 OSes like Unix) or read from the environment. Update the
4766 manpages and fix a few bugs.
4769 *) Add a few manpages for some of the openssl commands.
4772 *) Fix the -revoke option in ca. It was freeing up memory twice,
4773 leaking and not finding already revoked certificates.
4776 *) Extensive changes to support certificate auxiliary information.
4777 This involves the use of X509_CERT_AUX structure and X509_AUX
4778 functions. An X509_AUX function such as PEM_read_X509_AUX()
4779 can still read in a certificate file in the usual way but it
4780 will also read in any additional "auxiliary information". By
4781 doing things this way a fair degree of compatibility can be
4782 retained: existing certificates can have this information added
4783 using the new 'x509' options.
4785 Current auxiliary information includes an "alias" and some trust
4786 settings. The trust settings will ultimately be used in enhanced
4787 certificate chain verification routines: currently a certificate
4788 can only be trusted if it is self signed and then it is trusted
4792 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
4793 The problem was that one of the replacement routines had not been working
4794 since SSLeay releases. For now the offending routine has been replaced
4795 with non-optimised assembler. Even so, this now gives around 95%
4796 performance improvement for 1024 bit RSA signs.
4799 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
4800 handling. Most clients have the effective key size in bits equal to
4801 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
4802 A few however don't do this and instead use the size of the decrypted key
4803 to determine the RC2 key length and the AlgorithmIdentifier to determine
4804 the effective key length. In this case the effective key length can still
4805 be 40 bits but the key length can be 168 bits for example. This is fixed
4806 by manually forcing an RC2 key into the EVP_PKEY structure because the
4807 EVP code can't currently handle unusual RC2 key sizes: it always assumes
4808 the key length and effective key length are equal.
4811 *) Add a bunch of functions that should simplify the creation of
4812 X509_NAME structures. Now you should be able to do:
4813 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
4814 and have it automatically work out the correct field type and fill in
4815 the structures. The more adventurous can try:
4816 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
4817 and it will (hopefully) work out the correct multibyte encoding.
4820 *) Change the 'req' utility to use the new field handling and multibyte
4821 copy routines. Before the DN field creation was handled in an ad hoc
4822 way in req, ca, and x509 which was rather broken and didn't support
4823 BMPStrings or UTF8Strings. Since some software doesn't implement
4824 BMPStrings or UTF8Strings yet, they can be enabled using the config file
4825 using the dirstring_type option. See the new comment in the default
4826 openssl.cnf for more info.
4829 *) Make crypto/rand/md_rand.c more robust:
4830 - Assure unique random numbers after fork().
4831 - Make sure that concurrent threads access the global counter and
4832 md serializably so that we never lose entropy in them
4833 or use exactly the same state in multiple threads.
4834 Access to the large state is not always serializable because
4835 the additional locking could be a performance killer, and
4836 md should be large enough anyway.
4839 *) New file apps/app_rand.c with commonly needed functionality
4840 for handling the random seed file.
4842 Use the random seed file in some applications that previously did not:
4844 dsaparam -genkey (which also ignored its '-rand' option),
4847 x509 (when signing).
4848 Except on systems with /dev/urandom, it is crucial to have a random
4849 seed file at least for key creation, DSA signing, and for DH exchanges;
4850 for RSA signatures we could do without one.
4852 gendh and gendsa (unlike genrsa) used to read only the first byte
4853 of each file listed in the '-rand' option. The function as previously
4854 found in genrsa is now in app_rand.c and is used by all programs
4855 that support '-rand'.
4858 *) In RAND_write_file, use mode 0600 for creating files;
4859 don't just chmod when it may be too late.
4862 *) Report an error from X509_STORE_load_locations
4863 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
4866 *) New function ASN1_mbstring_copy() this copies a string in either
4867 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
4868 into an ASN1_STRING type. A mask of permissible types is passed
4869 and it chooses the "minimal" type to use or an error if not type
4873 *) Add function equivalents to the various macros in asn1.h. The old
4874 macros are retained with an M_ prefix. Code inside the library can
4875 use the M_ macros. External code (including the openssl utility)
4876 should *NOT* in order to be "shared library friendly".
4879 *) Add various functions that can check a certificate's extensions
4880 to see if it usable for various purposes such as SSL client,
4881 server or S/MIME and CAs of these types. This is currently
4882 VERY EXPERIMENTAL but will ultimately be used for certificate chain
4883 verification. Also added a -purpose flag to x509 utility to
4884 print out all the purposes.
4887 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
4891 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
4892 for, obtain and decode and extension and obtain its critical flag.
4893 This allows all the necessary extension code to be handled in a
4894 single function call.
4897 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
4898 platforms. See crypto/rc4/rc4_enc.c for further details.
4901 *) New -noout option to asn1parse. This causes no output to be produced
4902 its main use is when combined with -strparse and -out to extract data
4903 from a file (which may not be in ASN.1 format).
4906 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
4907 when producing the local key id.
4908 [Richard Levitte <levitte@stacken.kth.se>]
4910 *) New option -dhparam in s_server. This allows a DH parameter file to be
4911 stated explicitly. If it is not stated then it tries the first server
4912 certificate file. The previous behaviour hard coded the filename
4916 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
4917 a public key to be input or output. For example:
4918 openssl rsa -in key.pem -pubout -out pubkey.pem
4919 Also added necessary DSA public key functions to handle this.
4922 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
4923 in the message. This was handled by allowing
4924 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
4925 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
4927 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
4928 to the end of the strings whereas this didn't. This would cause problems
4929 if strings read with d2i_ASN1_bytes() were later modified.
4930 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
4932 *) Fix for base64 decode bug. When a base64 bio reads only one line of
4933 data and it contains EOF it will end up returning an error. This is
4934 caused by input 46 bytes long. The cause is due to the way base64
4935 BIOs find the start of base64 encoded data. They do this by trying a
4936 trial decode on each line until they find one that works. When they
4937 do a flag is set and it starts again knowing it can pass all the
4938 data directly through the decoder. Unfortunately it doesn't reset
4939 the context it uses. This means that if EOF is reached an attempt
4940 is made to pass two EOFs through the context and this causes the
4941 resulting error. This can also cause other problems as well. As is
4942 usual with these problems it takes *ages* to find and the fix is
4943 trivial: move one line.
4944 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
4946 *) Ugly workaround to get s_client and s_server working under Windows. The
4947 old code wouldn't work because it needed to select() on sockets and the
4948 tty (for keypresses and to see if data could be written). Win32 only
4949 supports select() on sockets so we select() with a 1s timeout on the
4950 sockets and then see if any characters are waiting to be read, if none
4951 are present then we retry, we also assume we can always write data to
4952 the tty. This isn't nice because the code then blocks until we've
4953 received a complete line of data and it is effectively polling the
4954 keyboard at 1s intervals: however it's quite a bit better than not
4955 working at all :-) A dedicated Windows application might handle this
4956 with an event loop for example.
4959 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
4960 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
4961 will be called when RSA_sign() and RSA_verify() are used. This is useful
4962 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
4963 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
4964 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
4965 This necessitated the support of an extra signature type NID_md5_sha1
4966 for SSL signatures and modifications to the SSL library to use it instead
4967 of calling RSA_public_decrypt() and RSA_private_encrypt().
4970 *) Add new -verify -CAfile and -CApath options to the crl program, these
4971 will lookup a CRL issuers certificate and verify the signature in a
4972 similar way to the verify program. Tidy up the crl program so it
4973 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
4974 less strict. It will now permit CRL extensions even if it is not
4975 a V2 CRL: this will allow it to tolerate some broken CRLs.
4978 *) Initialize all non-automatic variables each time one of the openssl
4979 sub-programs is started (this is necessary as they may be started
4980 multiple times from the "OpenSSL>" prompt).
4981 [Lennart Bang, Bodo Moeller]
4983 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
4984 removing all other RSA functionality (this is what NO_RSA does). This
4985 is so (for example) those in the US can disable those operations covered
4986 by the RSA patent while allowing storage and parsing of RSA keys and RSA
4990 *) Non-copying interface to BIO pairs.
4991 (still largely untested)
4994 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
4995 ASCII string. This was handled independently in various places before.
4998 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
4999 UTF8 strings a character at a time.
5002 *) Use client_version from client hello to select the protocol
5003 (s23_srvr.c) and for RSA client key exchange verification
5004 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
5007 *) Add various utility functions to handle SPKACs, these were previously
5008 handled by poking round in the structure internals. Added new function
5009 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
5010 print, verify and generate SPKACs. Based on an original idea from
5011 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
5014 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
5017 *) Allow the config file extension section to be overwritten on the
5018 command line. Based on an original idea from Massimiliano Pala
5019 <madwolf@comune.modena.it>. The new option is called -extensions
5020 and can be applied to ca, req and x509. Also -reqexts to override
5021 the request extensions in req and -crlexts to override the crl extensions
5025 *) Add new feature to the SPKAC handling in ca. Now you can include
5026 the same field multiple times by preceding it by "XXXX." for example:
5029 this is the same syntax as used in the req config file.
5032 *) Allow certificate extensions to be added to certificate requests. These
5033 are specified in a 'req_extensions' option of the req section of the
5034 config file. They can be printed out with the -text option to req but
5035 are otherwise ignored at present.
5038 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
5039 data read consists of only the final block it would not decrypted because
5040 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
5041 A misplaced 'break' also meant the decrypted final block might not be
5042 copied until the next read.
5045 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
5046 a few extra parameters to the DH structure: these will be useful if
5047 for example we want the value of 'q' or implement X9.42 DH.
5050 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
5051 provides hooks that allow the default DSA functions or functions on a
5052 "per key" basis to be replaced. This allows hardware acceleration and
5053 hardware key storage to be handled without major modification to the
5054 library. Also added low level modexp hooks and CRYPTO_EX structure and
5055 associated functions.
5058 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
5059 as "read only": it can't be written to and the buffer it points to will
5060 not be freed. Reading from a read only BIO is much more efficient than
5061 a normal memory BIO. This was added because there are several times when
5062 an area of memory needs to be read from a BIO. The previous method was
5063 to create a memory BIO and write the data to it, this results in two
5064 copies of the data and an O(n^2) reading algorithm. There is a new
5065 function BIO_new_mem_buf() which creates a read only memory BIO from
5066 an area of memory. Also modified the PKCS#7 routines to use read only
5070 *) Bugfix: ssl23_get_client_hello did not work properly when called in
5071 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
5072 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
5073 but a retry condition occured while trying to read the rest.
5076 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
5077 NID_pkcs7_encrypted by default: this was wrong since this should almost
5078 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
5079 the encrypted data type: this is a more sensible place to put it and it
5080 allows the PKCS#12 code to be tidied up that duplicated this
5084 *) Changed obj_dat.pl script so it takes its input and output files on
5085 the command line. This should avoid shell escape redirection problems
5089 *) Initial support for certificate extension requests, these are included
5090 in things like Xenroll certificate requests. Included functions to allow
5091 extensions to be obtained and added.
5094 *) -crlf option to s_client and s_server for sending newlines as
5095 CRLF (as required by many protocols).
5098 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
5100 *) Install libRSAglue.a when OpenSSL is built with RSAref.
5101 [Ralf S. Engelschall]
5103 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
5104 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
5106 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
5110 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
5111 DH parameters/keys (q is lost during that conversion, but the resulting
5112 DH parameters contain its length).
5114 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
5115 much faster than DH_generate_parameters (which creates parameters
5116 where p = 2*q + 1), and also the smaller q makes DH computations
5117 much more efficient (160-bit exponentiation instead of 1024-bit
5118 exponentiation); so this provides a convenient way to support DHE
5119 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
5120 utter importance to use
5121 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
5123 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
5124 when such DH parameters are used, because otherwise small subgroup
5125 attacks may become possible!
5128 *) Avoid memory leak in i2d_DHparams.
5131 *) Allow the -k option to be used more than once in the enc program:
5132 this allows the same encrypted message to be read by multiple recipients.
5135 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
5136 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
5137 it will always use the numerical form of the OID, even if it has a short
5141 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
5142 method only got called if p,q,dmp1,dmq1,iqmp components were present,
5143 otherwise bn_mod_exp was called. In the case of hardware keys for example
5144 no private key components need be present and it might store extra data
5145 in the RSA structure, which cannot be accessed from bn_mod_exp.
5146 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
5147 private key operations.
5150 *) Added support for SPARC Linux.
5153 *) pem_password_cb function type incompatibly changed from
5154 typedef int pem_password_cb(char *buf, int size, int rwflag);
5156 ....(char *buf, int size, int rwflag, void *userdata);
5157 so that applications can pass data to their callbacks:
5158 The PEM[_ASN1]_{read,write}... functions and macros now take an
5159 additional void * argument, which is just handed through whenever
5160 the password callback is called.
5161 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
5163 New function SSL_CTX_set_default_passwd_cb_userdata.
5165 Compatibility note: As many C implementations push function arguments
5166 onto the stack in reverse order, the new library version is likely to
5167 interoperate with programs that have been compiled with the old
5168 pem_password_cb definition (PEM_whatever takes some data that
5169 happens to be on the stack as its last argument, and the callback
5170 just ignores this garbage); but there is no guarantee whatsoever that
5173 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
5174 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
5175 problems not only on Windows, but also on some Unix platforms.
5176 To avoid problematic command lines, these definitions are now in an
5177 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
5178 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
5181 *) MIPS III/IV assembler module is reimplemented.
5184 *) More DES library cleanups: remove references to srand/rand and
5185 delete an unused file.
5188 *) Add support for the the free Netwide assembler (NASM) under Win32,
5189 since not many people have MASM (ml) and it can be hard to obtain.
5190 This is currently experimental but it seems to work OK and pass all
5191 the tests. Check out INSTALL.W32 for info.
5194 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
5195 without temporary keys kept an extra copy of the server key,
5196 and connections with temporary keys did not free everything in case
5200 *) New function RSA_check_key and new openssl rsa option -check
5201 for verifying the consistency of RSA keys.
5202 [Ulf Moeller, Bodo Moeller]
5204 *) Various changes to make Win32 compile work:
5205 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
5206 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
5207 comparison" warnings.
5208 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
5211 *) Add a debugging option to PKCS#5 v2 key generation function: when
5212 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
5213 derived keys are printed to stderr.
5216 *) Copy the flags in ASN1_STRING_dup().
5217 [Roman E. Pavlov <pre@mo.msk.ru>]
5219 *) The x509 application mishandled signing requests containing DSA
5220 keys when the signing key was also DSA and the parameters didn't match.
5222 It was supposed to omit the parameters when they matched the signing key:
5223 the verifying software was then supposed to automatically use the CA's
5224 parameters if they were absent from the end user certificate.
5226 Omitting parameters is no longer recommended. The test was also
5227 the wrong way round! This was probably due to unusual behaviour in
5228 EVP_cmp_parameters() which returns 1 if the parameters match.
5229 This meant that parameters were omitted when they *didn't* match and
5230 the certificate was useless. Certificates signed with 'ca' didn't have
5232 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
5234 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
5235 The interface is as follows:
5236 Applications can use
5237 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
5238 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
5239 "off" is now the default.
5240 The library internally uses
5241 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
5242 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
5243 to disable memory-checking temporarily.
5245 Some inconsistent states that previously were possible (and were
5246 even the default) are now avoided.
5248 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
5249 with each memory chunk allocated; this is occasionally more helpful
5250 than just having a counter.
5252 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
5254 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
5258 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
5259 which largely parallels "options", but is for changing API behaviour,
5260 whereas "options" are about protocol behaviour.
5261 Initial "mode" flags are:
5263 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
5264 a single record has been written.
5265 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
5266 retries use the same buffer location.
5267 (But all of the contents must be
5271 *) Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
5274 *) Fix problems with no-hmac etc.
5275 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
5277 *) New functions RSA_get_default_method(), RSA_set_method() and
5278 RSA_get_method(). These allows replacement of RSA_METHODs without having
5279 to mess around with the internals of an RSA structure.
5282 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
5283 Also really enable memory leak checks in openssl.c and in some
5285 [Chad C. Mulligan, Bodo Moeller]
5287 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
5288 up the length of negative integers. This has now been simplified to just
5289 store the length when it is first determined and use it later, rather
5290 than trying to keep track of where data is copied and updating it to
5292 [Steve Henson, reported by Brien Wheeler
5293 <bwheeler@authentica-security.com>]
5295 *) Add a new function PKCS7_signatureVerify. This allows the verification
5296 of a PKCS#7 signature but with the signing certificate passed to the
5297 function itself. This contrasts with PKCS7_dataVerify which assumes the
5298 certificate is present in the PKCS#7 structure. This isn't always the
5299 case: certificates can be omitted from a PKCS#7 structure and be
5300 distributed by "out of band" means (such as a certificate database).
5303 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
5304 function prototypes in pem.h, also change util/mkdef.pl to add the
5305 necessary function names.
5308 *) mk1mf.pl (used by Windows builds) did not properly read the
5309 options set by Configure in the top level Makefile, and Configure
5310 was not even able to write more than one option correctly.
5311 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
5314 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
5315 file to be loaded from a BIO or FILE pointer. The BIO version will
5316 for example allow memory BIOs to contain config info.
5319 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
5320 Whoever hopes to achieve shared-library compatibility across versions
5321 must use this, not the compile-time macro.
5322 (Exercise 0.9.4: Which is the minimum library version required by
5324 Note: All this applies only to multi-threaded programs, others don't
5328 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
5329 through a BIO pair triggered the default case, i.e.
5330 SSLerr(...,SSL_R_UNKNOWN_STATE).
5333 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
5334 can use the SSL library even if none of the specific BIOs is
5338 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
5339 for the encoded length.
5340 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
5342 *) Add initial documentation of the X509V3 functions.
5345 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
5346 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
5347 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
5348 secure PKCS#8 private key format with a high iteration count.
5351 *) Fix determination of Perl interpreter: A perl or perl5
5352 _directory_ in $PATH was also accepted as the interpreter.
5353 [Ralf S. Engelschall]
5355 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
5356 wrong with it but it was very old and did things like calling
5357 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
5361 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
5362 to use the new extension code.
5365 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
5366 with macros. This should make it easier to change their form, add extra
5367 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
5371 *) Add to configuration table a new entry that can specify an alternative
5372 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
5373 according to Mark Crispin <MRC@Panda.COM>.
5377 *) DES CBC did not update the IV. Weird.
5380 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
5381 Changing the behaviour of the former might break existing programs --
5382 where IV updating is needed, des_ncbc_encrypt can be used.
5385 *) When bntest is run from "make test" it drives bc to check its
5386 calculations, as well as internally checking them. If an internal check
5387 fails, it needs to cause bc to give a non-zero result or make test carries
5388 on without noticing the failure. Fixed.
5391 *) DES library cleanups.
5394 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
5395 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
5396 ciphers. NOTE: although the key derivation function has been verified
5397 against some published test vectors it has not been extensively tested
5398 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
5402 *) Instead of "mkdir -p", which is not fully portable, use new
5403 Perl script "util/mkdir-p.pl".
5406 *) Rewrite the way password based encryption (PBE) is handled. It used to
5407 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
5408 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
5409 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
5410 the 'parameter' field of the AlgorithmIdentifier is passed to the
5411 underlying key generation function so it must do its own ASN1 parsing.
5412 This has also changed the EVP_PBE_CipherInit() function which now has a
5413 'parameter' argument instead of literal salt and iteration count values
5414 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
5417 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
5418 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
5419 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
5420 KEY" because this clashed with PKCS#8 unencrypted string. Since this
5421 value was just used as a "magic string" and not used directly its
5422 value doesn't matter.
5425 *) Introduce some semblance of const correctness to BN. Shame C doesn't
5429 *) "linux-sparc64" configuration (ultrapenguin).
5430 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
5431 "linux-sparc" configuration.
5432 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
5434 *) config now generates no-xxx options for missing ciphers.
5437 *) Support the EBCDIC character set (work in progress).
5438 File ebcdic.c not yet included because it has a different license.
5439 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
5441 *) Support BS2000/OSD-POSIX.
5442 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
5444 *) Make callbacks for key generation use void * instead of char *.
5447 *) Make S/MIME samples compile (not yet tested).
5450 *) Additional typesafe stacks.
5453 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
5457 Changes between 0.9.3 and 0.9.3a [29 May 1999]
5459 *) New configuration variant "sco5-gcc".
5461 *) Updated some demos.
5462 [Sean O Riordain, Wade Scholine]
5464 *) Add missing BIO_free at exit of pkcs12 application.
5467 *) Fix memory leak in conf.c.
5470 *) Updates for Win32 to assembler version of MD5.
5473 *) Set #! path to perl in apps/der_chop to where we found it
5474 instead of using a fixed path.
5477 *) SHA library changes for irix64-mips4-cc.
5480 *) Improvements for VMS support.
5484 Changes between 0.9.2b and 0.9.3 [24 May 1999]
5486 *) Bignum library bug fix. IRIX 6 passes "make test" now!
5487 This also avoids the problems with SC4.2 and unpatched SC5.
5488 [Andy Polyakov <appro@fy.chalmers.se>]
5490 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
5491 These are required because of the typesafe stack would otherwise break
5492 existing code. If old code used a structure member which used to be STACK
5493 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
5494 sk_num or sk_value it would produce an error because the num, data members
5495 are not present in STACK_OF. Now it just produces a warning. sk_set
5496 replaces the old method of assigning a value to sk_value
5497 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
5498 that does this will no longer work (and should use sk_set instead) but
5499 this could be regarded as a "questionable" behaviour anyway.
5502 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
5503 correctly handle encrypted S/MIME data.
5506 *) Change type of various DES function arguments from des_cblock
5507 (which means, in function argument declarations, pointer to char)
5508 to des_cblock * (meaning pointer to array with 8 char elements),
5509 which allows the compiler to do more typechecking; it was like
5510 that back in SSLeay, but with lots of ugly casts.
5512 Introduce new type const_des_cblock.
5515 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
5516 problems: find RecipientInfo structure that matches recipient certificate
5517 and initialise the ASN1 structures properly based on passed cipher.
5520 *) Belatedly make the BN tests actually check the results.
5523 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
5524 to and from BNs: it was completely broken. New compilation option
5525 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
5526 key elements as negative integers.
5529 *) Reorganize and speed up MD5.
5530 [Andy Polyakov <appro@fy.chalmers.se>]
5533 [Richard Levitte <richard@levitte.org>]
5535 *) New option -out to asn1parse to allow the parsed structure to be
5536 output to a file. This is most useful when combined with the -strparse
5537 option to examine the output of things like OCTET STRINGS.
5540 *) Make SSL library a little more fool-proof by not requiring any longer
5541 that SSL_set_{accept,connect}_state be called before
5542 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
5543 in many applications because usually everything *appeared* to work as
5544 intended anyway -- now it really works as intended).
5547 *) Move openssl.cnf out of lib/.
5550 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
5551 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
5552 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
5553 [Ralf S. Engelschall]
5555 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
5556 handle PKCS#7 enveloped data properly.
5557 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
5559 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
5560 copying pointers. The cert_st handling is changed by this in
5561 various ways (and thus what used to be known as ctx->default_cert
5562 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
5563 any longer when s->cert does not give us what we need).
5564 ssl_cert_instantiate becomes obsolete by this change.
5565 As soon as we've got the new code right (possibly it already is?),
5566 we have solved a couple of bugs of the earlier code where s->cert
5567 was used as if it could not have been shared with other SSL structures.
5569 Note that using the SSL API in certain dirty ways now will result
5570 in different behaviour than observed with earlier library versions:
5571 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
5572 does not influence s as it used to.
5574 In order to clean up things more thoroughly, inside SSL_SESSION
5575 we don't use CERT any longer, but a new structure SESS_CERT
5576 that holds per-session data (if available); currently, this is
5577 the peer's certificate chain and, for clients, the server's certificate
5578 and temporary key. CERT holds only those values that can have
5579 meaningful defaults in an SSL_CTX.
5582 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
5583 from the internal representation. Various PKCS#7 fixes: remove some
5584 evil casts and set the enc_dig_alg field properly based on the signing
5588 *) Allow PKCS#12 password to be set from the command line or the
5589 environment. Let 'ca' get its config file name from the environment
5590 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
5594 *) Allow certificate policies extension to use an IA5STRING for the
5595 organization field. This is contrary to the PKIX definition but
5596 VeriSign uses it and IE5 only recognises this form. Document 'x509'
5600 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
5601 without disallowing inline assembler and the like for non-pedantic builds.
5604 *) Support Borland C++ builder.
5605 [Janez Jere <jj@void.si>, modified by Ulf Möller]
5610 *) SHA-1 cleanups and performance enhancements.
5611 [Andy Polyakov <appro@fy.chalmers.se>]
5613 *) Sparc v8plus assembler for the bignum library.
5614 [Andy Polyakov <appro@fy.chalmers.se>]
5616 *) Accept any -xxx and +xxx compiler options in Configure.
5619 *) Update HPUX configuration.
5622 *) Add missing sk_<type>_unshift() function to safestack.h
5623 [Ralf S. Engelschall]
5625 *) New function SSL_CTX_use_certificate_chain_file that sets the
5626 "extra_cert"s in addition to the certificate. (This makes sense
5627 only for "PEM" format files, as chains as a whole are not
5631 *) Support verify_depth from the SSL API.
5632 x509_vfy.c had what can be considered an off-by-one-error:
5633 Its depth (which was not part of the external interface)
5634 was actually counting the number of certificates in a chain;
5635 now it really counts the depth.
5638 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
5639 instead of X509err, which often resulted in confusing error
5640 messages since the error codes are not globally unique
5641 (e.g. an alleged error in ssl3_accept when a certificate
5642 didn't match the private key).
5644 *) New function SSL_CTX_set_session_id_context that allows to set a default
5645 value (so that you don't need SSL_set_session_id_context for each
5646 connection using the SSL_CTX).
5649 *) OAEP decoding bug fix.
5652 *) Support INSTALL_PREFIX for package builders, as proposed by
5656 *) New Configure options "threads" and "no-threads". For systems
5657 where the proper compiler options are known (currently Solaris
5658 and Linux), "threads" is the default.
5661 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
5664 *) Install various scripts to $(OPENSSLDIR)/misc, not to
5665 $(INSTALLTOP)/bin -- they shouldn't clutter directories
5666 such as /usr/local/bin.
5669 *) "make linux-shared" to build shared libraries.
5670 [Niels Poppe <niels@netbox.org>]
5672 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
5675 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
5676 extension adding in x509 utility.
5679 *) Remove NOPROTO sections and error code comments.
5682 *) Partial rewrite of the DEF file generator to now parse the ANSI
5686 *) New Configure options --prefix=DIR and --openssldir=DIR.
5689 *) Complete rewrite of the error code script(s). It is all now handled
5690 by one script at the top level which handles error code gathering,
5691 header rewriting and C source file generation. It should be much better
5692 than the old method: it now uses a modified version of Ulf's parser to
5693 read the ANSI prototypes in all header files (thus the old K&R definitions
5694 aren't needed for error creation any more) and do a better job of
5695 translating function codes into names. The old 'ASN1 error code imbedded
5696 in a comment' is no longer necessary and it doesn't use .err files which
5697 have now been deleted. Also the error code call doesn't have to appear all
5698 on one line (which resulted in some large lines...).
5701 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
5704 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
5705 0 (which usually indicates a closed connection), but continue reading.
5708 *) Fix some race conditions.
5711 *) Add support for CRL distribution points extension. Add Certificate
5712 Policies and CRL distribution points documentation.
5715 *) Move the autogenerated header file parts to crypto/opensslconf.h.
5718 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
5719 8 of keying material. Merlin has also confirmed interop with this fix
5720 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
5721 [Merlin Hughes <merlin@baltimore.ie>]
5723 *) Fix lots of warnings.
5724 [Richard Levitte <levitte@stacken.kth.se>]
5726 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
5727 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
5728 [Richard Levitte <levitte@stacken.kth.se>]
5730 *) Fix problems with sizeof(long) == 8.
5731 [Andy Polyakov <appro@fy.chalmers.se>]
5733 *) Change functions to ANSI C.
5736 *) Fix typos in error codes.
5737 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
5739 *) Remove defunct assembler files from Configure.
5742 *) SPARC v8 assembler BIGNUM implementation.
5743 [Andy Polyakov <appro@fy.chalmers.se>]
5745 *) Support for Certificate Policies extension: both print and set.
5746 Various additions to support the r2i method this uses.
5749 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
5750 return a const string when you are expecting an allocated buffer.
5753 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
5754 types DirectoryString and DisplayText.
5757 *) Add code to allow r2i extensions to access the configuration database,
5758 add an LHASH database driver and add several ctx helper functions.
5761 *) Fix an evil bug in bn_expand2() which caused various BN functions to
5762 fail when they extended the size of a BIGNUM.
5765 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
5766 support typesafe stack.
5769 *) Fix typo in SSL_[gs]et_options().
5770 [Nils Frostberg <nils@medcom.se>]
5772 *) Delete various functions and files that belonged to the (now obsolete)
5773 old X509V3 handling code.
5776 *) New Configure option "rsaref".
5779 *) Don't auto-generate pem.h.
5782 *) Introduce type-safe ASN.1 SETs.
5785 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
5786 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
5788 *) Introduce type-safe STACKs. This will almost certainly break lots of code
5789 that links with OpenSSL (well at least cause lots of warnings), but fear
5790 not: the conversion is trivial, and it eliminates loads of evil casts. A
5791 few STACKed things have been converted already. Feel free to convert more.
5792 In the fullness of time, I'll do away with the STACK type altogether.
5795 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
5796 specified in <certfile> by updating the entry in the index.txt file.
5797 This way one no longer has to edit the index.txt file manually for
5798 revoking a certificate. The -revoke option does the gory details now.
5799 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
5801 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
5802 `-text' option at all and this way the `-noout -text' combination was
5803 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
5804 [Ralf S. Engelschall]
5806 *) Make sure a corresponding plain text error message exists for the
5807 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
5808 verify callback function determined that a certificate was revoked.
5809 [Ralf S. Engelschall]
5811 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
5812 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
5813 all available cipers including rc5, which was forgotten until now.
5814 In order to let the testing shell script know which algorithms
5815 are available, a new (up to now undocumented) command
5816 "openssl list-cipher-commands" is used.
5819 *) Bugfix: s_client occasionally would sleep in select() when
5820 it should have checked SSL_pending() first.
5823 *) New functions DSA_do_sign and DSA_do_verify to provide access to
5824 the raw DSA values prior to ASN.1 encoding.
5827 *) Tweaks to Configure
5828 [Niels Poppe <niels@netbox.org>]
5830 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
5834 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
5837 *) New config option to avoid instructions that are illegal on the 80386.
5838 The default code is faster, but requires at least a 486.
5841 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
5842 SSL2_SERVER_VERSION (not used at all) macros, which are now the
5843 same as SSL2_VERSION anyway.
5846 *) New "-showcerts" option for s_client.
5849 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
5850 application. Various cleanups and fixes.
5853 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
5854 modify error routines to work internally. Add error codes and PBE init
5855 to library startup routines.
5858 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
5859 packing functions to asn1 and evp. Changed function names and error
5860 codes along the way.
5863 *) PKCS12 integration: and so it begins... First of several patches to
5864 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
5865 objects to objects.h
5868 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
5869 and display support for Thawte strong extranet extension.
5872 *) Add LinuxPPC support.
5873 [Jeff Dubrule <igor@pobox.org>]
5875 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
5876 bn_div_words in alpha.s.
5877 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
5879 *) Make sure the RSA OAEP test is skipped under -DRSAref because
5880 OAEP isn't supported when OpenSSL is built with RSAref.
5881 [Ulf Moeller <ulf@fitug.de>]
5883 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
5884 so they no longer are missing under -DNOPROTO.
5885 [Soren S. Jorvang <soren@t.dk>]
5888 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
5890 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
5891 doesn't work when the session is reused. Coming soon!
5894 *) Fix a security hole, that allows sessions to be reused in the wrong
5895 context thus bypassing client cert protection! All software that uses
5896 client certs and session caches in multiple contexts NEEDS PATCHING to
5897 allow session reuse! A fuller solution is in the works.
5898 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
5900 *) Some more source tree cleanups (removed obsolete files
5901 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
5902 permission on "config" script to be executable) and a fix for the INSTALL
5904 [Ulf Moeller <ulf@fitug.de>]
5906 *) Remove some legacy and erroneous uses of malloc, free instead of
5908 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
5910 *) Make rsa_oaep_test return non-zero on error.
5911 [Ulf Moeller <ulf@fitug.de>]
5913 *) Add support for native Solaris shared libraries. Configure
5914 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
5915 if someone would make that last step automatic.
5916 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
5918 *) ctx_size was not built with the right compiler during "make links". Fixed.
5921 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
5922 except NULL ciphers". This means the default cipher list will no longer
5923 enable NULL ciphers. They need to be specifically enabled e.g. with
5924 the string "DEFAULT:eNULL".
5927 *) Fix to RSA private encryption routines: if p < q then it would
5928 occasionally produce an invalid result. This will only happen with
5929 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
5932 *) Be less restrictive and allow also `perl util/perlpath.pl
5933 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
5934 because this way one can also use an interpreter named `perl5' (which is
5935 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
5936 installed as `perl').
5937 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
5939 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
5940 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
5942 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
5943 advapi32.lib to Win32 build and change the pem test comparision
5944 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
5945 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
5946 and crypto/des/ede_cbcm_enc.c.
5949 *) DES quad checksum was broken on big-endian architectures. Fixed.
5952 *) Comment out two functions in bio.h that aren't implemented. Fix up the
5953 Win32 test batch file so it (might) work again. The Win32 test batch file
5954 is horrible: I feel ill....
5957 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
5958 in e_os.h. Audit of header files to check ANSI and non ANSI
5959 sections: 10 functions were absent from non ANSI section and not exported
5960 from Windows DLLs. Fixed up libeay.num for new functions.
5963 *) Make `openssl version' output lines consistent.
5964 [Ralf S. Engelschall]
5966 *) Fix Win32 symbol export lists for BIO functions: Added
5967 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
5968 to ms/libeay{16,32}.def.
5969 [Ralf S. Engelschall]
5971 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
5972 fine under Unix and passes some trivial tests I've now added. But the
5973 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
5974 added to make sure no one expects that this stuff really works in the
5975 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
5976 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
5978 [Ralf S. Engelschall]
5980 *) Fix the generation of two part addresses in perl.
5981 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
5983 *) Add config entry for Linux on MIPS.
5984 [John Tobey <jtobey@channel1.com>]
5986 *) Make links whenever Configure is run, unless we are on Windoze.
5989 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
5990 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
5994 *) Add a useful kludge to allow package maintainers to specify compiler and
5995 other platforms details on the command line without having to patch the
5996 Configure script everytime: One now can use ``perl Configure
5997 <id>:<details>'', i.e. platform ids are allowed to have details appended
5998 to them (seperated by colons). This is treated as there would be a static
5999 pre-configured entry in Configure's %table under key <id> with value
6000 <details> and ``perl Configure <id>'' is called. So, when you want to
6001 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
6002 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
6003 now, which overrides the FreeBSD-elf entry on-the-fly.
6004 [Ralf S. Engelschall]
6006 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
6009 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
6010 on the `perl Configure ...' command line. This way one can compile
6011 OpenSSL libraries with Position Independent Code (PIC) which is needed
6012 for linking it into DSOs.
6013 [Ralf S. Engelschall]
6015 *) Remarkably, export ciphers were totally broken and no-one had noticed!
6019 *) Cleaned up the LICENSE document: The official contact for any license
6020 questions now is the OpenSSL core team under openssl-core@openssl.org.
6021 And add a paragraph about the dual-license situation to make sure people
6022 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
6023 to the OpenSSL toolkit.
6024 [Ralf S. Engelschall]
6026 *) General source tree makefile cleanups: Made `making xxx in yyy...'
6027 display consistent in the source tree and replaced `/bin/rm' by `rm'.
6028 Additonally cleaned up the `make links' target: Remove unnecessary
6029 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
6030 to speed processing and no longer clutter the display with confusing
6031 stuff. Instead only the actually done links are displayed.
6032 [Ralf S. Engelschall]
6034 *) Permit null encryption ciphersuites, used for authentication only. It used
6035 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
6036 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
6040 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
6041 signed attributes when verifying signatures (this would break them),
6042 the detached data encoding was wrong and public keys obtained using
6043 X509_get_pubkey() weren't freed.
6046 *) Add text documentation for the BUFFER functions. Also added a work around
6047 to a Win95 console bug. This was triggered by the password read stuff: the
6048 last character typed gets carried over to the next fread(). If you were
6049 generating a new cert request using 'req' for example then the last
6050 character of the passphrase would be CR which would then enter the first
6054 *) Added the new `Includes OpenSSL Cryptography Software' button as
6055 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
6056 button and can be used by applications based on OpenSSL to show the
6057 relationship to the OpenSSL project.
6058 [Ralf S. Engelschall]
6060 *) Remove confusing variables in function signatures in files
6061 ssl/ssl_lib.c and ssl/ssl.h.
6062 [Lennart Bong <lob@kulthea.stacken.kth.se>]
6064 *) Don't install bss_file.c under PREFIX/include/
6065 [Lennart Bong <lob@kulthea.stacken.kth.se>]
6067 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
6068 functions that return function pointers and has support for NT specific
6069 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
6070 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
6071 unsigned to signed types: this was killing the Win32 compile.
6074 *) Add new certificate file to stack functions,
6075 SSL_add_dir_cert_subjects_to_stack() and
6076 SSL_add_file_cert_subjects_to_stack(). These largely supplant
6077 SSL_load_client_CA_file(), and can be used to add multiple certs easily
6078 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
6079 This means that Apache-SSL and similar packages don't have to mess around
6080 to add as many CAs as they want to the preferred list.
6083 *) Experiment with doxygen documentation. Currently only partially applied to
6085 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
6086 openssl.doxy as the configuration file.
6089 *) Get rid of remaining C++-style comments which strict C compilers hate.
6090 [Ralf S. Engelschall, pointed out by Carlos Amengual]
6092 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
6093 compiled in by default: it has problems with large keys.
6096 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
6097 DH private keys and/or callback functions which directly correspond to
6098 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
6099 is needed for applications which have to configure certificates on a
6100 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
6102 For the RSA certificate situation is makes no difference, but
6103 for the DSA certificate situation this fixes the "no shared cipher"
6104 problem where the OpenSSL cipher selection procedure failed because the
6105 temporary keys were not overtaken from the context and the API provided
6106 no way to reconfigure them.
6107 The new functions now let applications reconfigure the stuff and they
6108 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
6109 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
6110 non-public-API function ssl_cert_instantiate() is used as a helper
6111 function and also to reduce code redundancy inside ssl_rsa.c.
6112 [Ralf S. Engelschall]
6114 *) Move s_server -dcert and -dkey options out of the undocumented feature
6115 area because they are useful for the DSA situation and should be
6116 recognized by the users.
6117 [Ralf S. Engelschall]
6119 *) Fix the cipher decision scheme for export ciphers: the export bits are
6120 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
6121 SSL_EXP_MASK. So, the original variable has to be used instead of the
6122 already masked variable.
6123 [Richard Levitte <levitte@stacken.kth.se>]
6125 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
6126 [Richard Levitte <levitte@stacken.kth.se>]
6128 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
6129 from `int' to `unsigned int' because it's a length and initialized by
6130 EVP_DigestFinal() which expects an `unsigned int *'.
6131 [Richard Levitte <levitte@stacken.kth.se>]
6133 *) Don't hard-code path to Perl interpreter on shebang line of Configure
6134 script. Instead use the usual Shell->Perl transition trick.
6135 [Ralf S. Engelschall]
6137 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
6138 (in addition to RSA certificates) to match the behaviour of `openssl dsa
6139 -noout -modulus' as it's already the case for `openssl rsa -noout
6140 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
6141 currently the public key is printed (a decision which was already done by
6142 `openssl dsa -modulus' in the past) which serves a similar purpose.
6143 Additionally the NO_RSA no longer completely removes the whole -modulus
6144 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
6146 [Ralf S. Engelschall]
6148 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
6149 BIO. See the source (crypto/evp/bio_ok.c) for more info.
6150 [Arne Ansper <arne@ats.cyber.ee>]
6152 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
6153 to be added. Now both 'req' and 'ca' can use new objects defined in the
6157 *) Add cool BIO that does syslog (or event log on NT).
6158 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
6160 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
6161 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
6162 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
6163 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
6166 *) Add preliminary config info for new extension code.
6169 *) Make RSA_NO_PADDING really use no padding.
6170 [Ulf Moeller <ulf@fitug.de>]
6172 *) Generate errors when private/public key check is done.
6175 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
6176 for some CRL extensions and new objects added.
6179 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
6180 key usage extension and fuller support for authority key id.
6183 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
6184 padding method for RSA, which is recommended for new applications in PKCS
6185 #1 v2.0 (RFC 2437, October 1998).
6186 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
6187 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
6188 against Bleichbacher's attack on RSA.
6189 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
6192 *) Updates to the new SSL compression code
6193 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
6195 *) Fix so that the version number in the master secret, when passed
6196 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
6197 (because the server will not accept higher), that the version number
6198 is 0x03,0x01, not 0x03,0x00
6199 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
6201 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
6202 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
6203 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
6206 *) Support for RAW extensions where an arbitrary extension can be
6207 created by including its DER encoding. See apps/openssl.cnf for
6211 *) Make sure latest Perl versions don't interpret some generated C array
6212 code as Perl array code in the crypto/err/err_genc.pl script.
6213 [Lars Weber <3weber@informatik.uni-hamburg.de>]
6215 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
6216 not many people have the assembler. Various Win32 compilation fixes and
6217 update to the INSTALL.W32 file with (hopefully) more accurate Win32
6221 *) Modify configure script 'Configure' to automatically create crypto/date.h
6222 file under Win32 and also build pem.h from pem.org. New script
6223 util/mkfiles.pl to create the MINFO file on environments that can't do a
6224 'make files': perl util/mkfiles.pl >MINFO should work.
6227 *) Major rework of DES function declarations, in the pursuit of correctness
6228 and purity. As a result, many evil casts evaporated, and some weirdness,
6229 too. You may find this causes warnings in your code. Zapping your evil
6230 casts will probably fix them. Mostly.
6233 *) Fix for a typo in asn1.h. Bug fix to object creation script
6234 obj_dat.pl. It considered a zero in an object definition to mean
6235 "end of object": none of the objects in objects.h have any zeros
6236 so it wasn't spotted.
6237 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
6239 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
6240 Masking (CBCM). In the absence of test vectors, the best I have been able
6241 to do is check that the decrypt undoes the encrypt, so far. Send me test
6242 vectors if you have them.
6245 *) Correct calculation of key length for export ciphers (too much space was
6246 allocated for null ciphers). This has not been tested!
6249 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
6250 message is now correct (it understands "crypto" and "ssl" on its
6251 command line). There is also now an "update" option. This will update
6252 the util/ssleay.num and util/libeay.num files with any new functions.
6254 perl util/mkdef.pl crypto ssl update
6255 it will update them.
6258 *) Overhauled the Perl interface (perl/*):
6259 - ported BN stuff to OpenSSL's different BN library
6260 - made the perl/ source tree CVS-aware
6261 - renamed the package from SSLeay to OpenSSL (the files still contain
6262 their history because I've copied them in the repository)
6263 - removed obsolete files (the test scripts will be replaced
6264 by better Test::Harness variants in the future)
6265 [Ralf S. Engelschall]
6267 *) First cut for a very conservative source tree cleanup:
6268 1. merge various obsolete readme texts into doc/ssleay.txt
6269 where we collect the old documents and readme texts.
6270 2. remove the first part of files where I'm already sure that we no
6271 longer need them because of three reasons: either they are just temporary
6272 files which were left by Eric or they are preserved original files where
6273 I've verified that the diff is also available in the CVS via "cvs diff
6274 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
6275 the crypto/md/ stuff).
6276 [Ralf S. Engelschall]
6278 *) More extension code. Incomplete support for subject and issuer alt
6279 name, issuer and authority key id. Change the i2v function parameters
6280 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
6281 what that's for :-) Fix to ASN1 macro which messed up
6282 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
6285 *) Preliminary support for ENUMERATED type. This is largely copied from the
6289 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
6290 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
6292 *) Make sure `make rehash' target really finds the `openssl' program.
6293 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
6295 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
6296 like to hear about it if this slows down other processors.
6299 *) Add CygWin32 platform information to Configure script.
6300 [Alan Batie <batie@aahz.jf.intel.com>]
6302 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
6303 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
6305 *) New program nseq to manipulate netscape certificate sequences
6308 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
6312 *) Fixes to BN code. Previously the default was to define BN_RECURSION
6313 but the BN code had some problems that would cause failures when
6314 doing certificate verification and some other functions.
6315 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
6317 *) Add ASN1 and PEM code to support netscape certificate sequences.
6320 *) Add ASN1 and PEM code to support netscape certificate sequences.
6323 *) Add several PKIX and private extended key usage OIDs.
6326 *) Modify the 'ca' program to handle the new extension code. Modify
6327 openssl.cnf for new extension format, add comments.
6330 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
6331 and add a sample to openssl.cnf so req -x509 now adds appropriate
6335 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
6336 error code, add initial support to X509_print() and x509 application.
6339 *) Takes a deep breath and start addding X509 V3 extension support code. Add
6340 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
6341 stuff is currently isolated and isn't even compiled yet.
6344 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
6345 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
6346 Removed the versions check from X509 routines when loading extensions:
6347 this allows certain broken certificates that don't set the version
6348 properly to be processed.
6351 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
6352 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
6353 can still be regenerated with "make depend".
6356 *) Spelling mistake in C version of CAST-128.
6357 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
6359 *) Changes to the error generation code. The perl script err-code.pl
6360 now reads in the old error codes and retains the old numbers, only
6361 adding new ones if necessary. It also only changes the .err files if new
6362 codes are added. The makefiles have been modified to only insert errors
6363 when needed (to avoid needlessly modifying header files). This is done
6364 by only inserting errors if the .err file is newer than the auto generated
6365 C file. To rebuild all the error codes from scratch (the old behaviour)
6366 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
6367 or delete all the .err files.
6370 *) CAST-128 was incorrectly implemented for short keys. The C version has
6371 been fixed, but is untested. The assembler versions are also fixed, but
6372 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
6373 to regenerate it if needed.
6374 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
6375 Hagino <itojun@kame.net>]
6377 *) File was opened incorrectly in randfile.c.
6378 [Ulf Möller <ulf@fitug.de>]
6380 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
6381 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
6382 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
6383 al: it's just almost always a UTCTime. Note this patch adds new error
6384 codes so do a "make errors" if there are problems.
6387 *) Correct Linux 1 recognition in config.
6388 [Ulf Möller <ulf@fitug.de>]
6390 *) Remove pointless MD5 hash when using DSA keys in ca.
6391 [Anonymous <nobody@replay.com>]
6393 *) Generate an error if given an empty string as a cert directory. Also
6394 generate an error if handed NULL (previously returned 0 to indicate an
6395 error, but didn't set one).
6396 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
6398 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
6401 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
6402 parameters. This was causing a warning which killed off the Win32 compile.
6405 *) Remove C++ style comments from crypto/bn/bn_local.h.
6406 [Neil Costigan <neil.costigan@celocom.com>]
6408 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
6409 based on a text string, looking up short and long names and finally
6410 "dot" format. The "dot" format stuff didn't work. Added new function
6411 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
6412 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
6413 OID is not part of the table.
6416 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
6417 X509_LOOKUP_by_alias().
6420 *) Sort openssl functions by name.
6423 *) Get the gendsa program working (hopefully) and add it to app list. Remove
6424 encryption from sample DSA keys (in case anyone is interested the password
6428 *) Make _all_ *_free functions accept a NULL pointer.
6429 [Frans Heymans <fheymans@isaserver.be>]
6431 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
6433 [Anonymous <nobody@replay.com>]
6435 *) s_server should send the CAfile as acceptable CAs, not its own cert.
6436 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
6438 *) Don't blow it for numeric -newkey arguments to apps/req.
6439 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
6441 *) Temp key "for export" tests were wrong in s3_srvr.c.
6442 [Anonymous <nobody@replay.com>]
6444 *) Add prototype for temp key callback functions
6445 SSL_CTX_set_tmp_{rsa,dh}_callback().
6448 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
6449 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
6452 *) X509_name_add_entry() freed the wrong thing after an error.
6453 [Arne Ansper <arne@ats.cyber.ee>]
6455 *) rsa_eay.c would attempt to free a NULL context.
6456 [Arne Ansper <arne@ats.cyber.ee>]
6458 *) BIO_s_socket() had a broken should_retry() on Windoze.
6459 [Arne Ansper <arne@ats.cyber.ee>]
6461 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
6462 [Arne Ansper <arne@ats.cyber.ee>]
6464 *) Make sure the already existing X509_STORE->depth variable is initialized
6465 in X509_STORE_new(), but document the fact that this variable is still
6466 unused in the certificate verification process.
6467 [Ralf S. Engelschall]
6469 *) Fix the various library and apps files to free up pkeys obtained from
6470 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
6473 *) Fix reference counting in X509_PUBKEY_get(). This makes
6474 demos/maurice/example2.c work, amongst others, probably.
6475 [Steve Henson and Ben Laurie]
6477 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
6478 `openssl' and second, the shortcut symlinks for the `openssl <command>'
6479 are no longer created. This way we have a single and consistent command
6480 line interface `openssl <command>', similar to `cvs <command>'.
6481 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
6483 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
6484 BIT STRING wrapper always have zero unused bits.
6487 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
6490 *) Make the top-level INSTALL documentation easier to understand.
6493 *) Makefiles updated to exit if an error occurs in a sub-directory
6494 make (including if user presses ^C) [Paul Sutton]
6496 *) Make Montgomery context stuff explicit in RSA data structure.
6499 *) Fix build order of pem and err to allow for generated pem.h.
6502 *) Fix renumbering bug in X509_NAME_delete_entry().
6505 *) Enhanced the err-ins.pl script so it makes the error library number
6506 global and can add a library name. This is needed for external ASN1 and
6507 other error libraries.
6510 *) Fixed sk_insert which never worked properly.
6513 *) Fix ASN1 macros so they can handle indefinite length construted
6514 EXPLICIT tags. Some non standard certificates use these: they can now
6518 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
6519 into a single doc/ssleay.txt bundle. This way the information is still
6520 preserved but no longer messes up this directory. Now it's new room for
6521 the new set of documenation files.
6522 [Ralf S. Engelschall]
6524 *) SETs were incorrectly DER encoded. This was a major pain, because they
6525 shared code with SEQUENCEs, which aren't coded the same. This means that
6526 almost everything to do with SETs or SEQUENCEs has either changed name or
6527 number of arguments.
6528 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
6530 *) Fix test data to work with the above.
6533 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
6534 was already fixed by Eric for 0.9.1 it seems.
6535 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
6537 *) Autodetect FreeBSD3.
6540 *) Fix various bugs in Configure. This affects the following platforms:
6544 unixware-2.0-pentium
6548 *) Eliminate generated files from CVS. Reorder tests to regenerate files
6549 before they are needed.
6552 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
6556 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
6558 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
6559 changed SSLeay to OpenSSL in version strings.
6560 [Ralf S. Engelschall]
6562 *) Some fixups to the top-level documents.
6565 *) Fixed the nasty bug where rsaref.h was not found under compile-time
6566 because the symlink to include/ was missing.
6567 [Ralf S. Engelschall]
6569 *) Incorporated the popular no-RSA/DSA-only patches
6570 which allow to compile a RSA-free SSLeay.
6571 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
6573 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
6574 when "ssleay" is still not found.
6575 [Ralf S. Engelschall]
6577 *) Added more platforms to Configure: Cray T3E, HPUX 11,
6578 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
6580 *) Updated the README file.
6581 [Ralf S. Engelschall]
6583 *) Added various .cvsignore files in the CVS repository subdirs
6584 to make a "cvs update" really silent.
6585 [Ralf S. Engelschall]
6587 *) Recompiled the error-definition header files and added
6588 missing symbols to the Win32 linker tables.
6589 [Ralf S. Engelschall]
6591 *) Cleaned up the top-level documents;
6592 o new files: CHANGES and LICENSE
6593 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
6594 o merged COPYRIGHT into LICENSE
6595 o removed obsolete TODO file
6596 o renamed MICROSOFT to INSTALL.W32
6597 [Ralf S. Engelschall]
6599 *) Removed dummy files from the 0.9.1b source tree:
6600 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
6601 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
6602 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
6603 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
6604 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
6605 [Ralf S. Engelschall]
6607 *) Added various platform portability fixes.
6610 *) The Genesis of the OpenSSL rpject:
6611 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
6612 Young and Tim J. Hudson created while they were working for C2Net until
6614 [The OpenSSL Project]
6617 Changes between 0.9.0b and 0.9.1b [not released]
6619 *) Updated a few CA certificates under certs/
6622 *) Changed some BIGNUM api stuff.
6625 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
6626 DGUX x86, Linux Alpha, etc.
6629 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
6630 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
6634 *) Add -strparse option to asn1pars program which parses nested
6636 [Dr Stephen Henson <shenson@bigfoot.com>]
6638 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
6641 *) DSA fix for "ca" program.
6644 *) Added "-genkey" option to "dsaparam" program.
6647 *) Added RIPE MD160 (rmd160) message digest.
6650 *) Added -a (all) option to "ssleay version" command.
6653 *) Added PLATFORM define which is the id given to Configure.
6656 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
6659 *) Extended the ASN.1 parser routines.
6662 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
6665 *) Added a BN_CTX to the BN library.
6668 *) Fixed the weak key values in DES library
6671 *) Changed API in EVP library for cipher aliases.
6674 *) Added support for RC2/64bit cipher.
6677 *) Converted the lhash library to the crypto/mem.c functions.
6680 *) Added more recognized ASN.1 object ids.
6683 *) Added more RSA padding checks for SSL/TLS.
6686 *) Added BIO proxy/filter functionality.
6689 *) Added extra_certs to SSL_CTX which can be used
6690 send extra CA certificates to the client in the CA cert chain sending
6691 process. It can be configured with SSL_CTX_add_extra_chain_cert().
6694 *) Now Fortezza is denied in the authentication phase because
6695 this is key exchange mechanism is not supported by SSLeay at all.
6698 *) Additional PKCS1 checks.
6701 *) Support the string "TLSv1" for all TLS v1 ciphers.
6704 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
6705 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
6708 *) Fixed a few memory leaks.
6711 *) Fixed various code and comment typos.
6714 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
6715 bytes sent in the client random.
6716 [Edward Bishop <ebishop@spyglass.com>]