]>
Commit | Line | Data |
---|---|---|
1 | ||
2 | OpenSSL CHANGES | |
3 | _______________ | |
4 | ||
5 | Changes between 0.9.4 and 0.9.5 [xx XXX 1999] | |
6 | ||
7 | *) Support for ASN1 "NULL" type. This could be handled before by using | |
8 | ASN1_TYPE but there wasn't any function that would try to read a NULL | |
9 | and produce an error if it couldn't. For compatibility we also have | |
10 | ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and | |
11 | don't allocate anything because they don't need to. | |
12 | [Steve Henson] | |
13 | ||
14 | *) Initial support for MacOS is now provided. Examine INSTALL.MacOS | |
15 | for details. | |
16 | [Andy Polyakov, Roy Woods <roy@centicsystems.ca>] | |
17 | ||
18 | *) Rebuild of the memory allocation routines used by OpenSSL code and | |
19 | possibly others as well. The purpose is to make an interface that | |
20 | provide hooks so anyone can build a separate set of allocation and | |
21 | deallocation routines to be used by OpenSSL, for example if memory | |
22 | pool implementations, or something else. The same is provided for | |
23 | memory debugging code. OpenSSL already comes with code that finds | |
24 | memory leaks, but this gives people a chance to debug other memory | |
25 | problems. | |
26 | ||
27 | With these changes, a new set of functions and macros have appeared: | |
28 | ||
29 | CRYPTO_set_mem_debug_functions() [F] | |
30 | CRYPTO_get_mem_debug_functions() [F] | |
31 | CRYPTO_dbg_set_options() [F] | |
32 | CRYPTO_dbg_get_options() [F] | |
33 | CRYPTO_melloc_debug_init() [M] | |
34 | ||
35 | The memory debug functions are NULL by default, unless the library | |
36 | is compiled with CRYPTO_MDEBUG or friends is defined. If someone | |
37 | wants to debug memory anyway, CRYPTO_malloc_debug_init() or | |
38 | CRYPTO_set_mem_debug_functions() must be used. | |
39 | ||
40 | Also, things like CRYPTO_set_mem_functions will always give the | |
41 | expected result (the new set of functions is used for allocation | |
42 | and deallocation) at all times, regardless of platform and compiler | |
43 | options. | |
44 | ||
45 | To finish it up, some functions that were never use in any other | |
46 | way than through macros have a new API and new semantic: | |
47 | ||
48 | CRYPTO_dbg_malloc() | |
49 | CRYPTO_dbg_realloc() | |
50 | CRYPTO_dbg_free() | |
51 | ||
52 | All macros of value have retained their old syntax. | |
53 | [Richard Levitte] | |
54 | ||
55 | *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the | |
56 | ordering of SMIMECapabilities wasn't in "strength order" and there | |
57 | was a missing NULL in the AlgorithmIdentifier for the SHA1 signature | |
58 | algorithm. | |
59 | [Steve Henson] | |
60 | ||
61 | *) Some ASN1 types with illegal zero length encoding (INTEGER, | |
62 | ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines. | |
63 | [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson] | |
64 | ||
65 | *) Merge in my S/MIME library for OpenSSL. This provides a simple | |
66 | S/MIME API on top of the PKCS#7 code, a MIME parser (with enough | |
67 | functionality to handle multipart/signed properly) and a utility | |
68 | called 'smime' to call all this stuff. This is based on code I | |
69 | originally wrote for Celo who have kindly allowed it to be | |
70 | included in OpenSSL. | |
71 | [Steve Henson] | |
72 | ||
73 | *) Add variants des_set_key_checked and des_set_key_unchecked of | |
74 | des_set_key (aka des_key_sched). Global variable des_check_key | |
75 | decides which of these is called by des_set_key; this way | |
76 | des_check_key behaves as it always did, but applications and | |
77 | the library itself, which was buggy for des_check_key == 1, | |
78 | have a cleaner way to pick the version they need. | |
79 | [Bodo Moeller] | |
80 | ||
81 | *) New function PKCS12_newpass() which changes the password of a | |
82 | PKCS12 structure. | |
83 | [Steve Henson] | |
84 | ||
85 | *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and | |
86 | dynamic mix. In both cases the ids can be used as an index into the | |
87 | table. Also modified the X509_TRUST_add() and X509_PURPOSE_add() | |
88 | functions so they accept a list of the field values and the | |
89 | application doesn't need to directly manipulate the X509_TRUST | |
90 | structure. | |
91 | [Steve Henson] | |
92 | ||
93 | *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't | |
94 | need initialising. | |
95 | [Steve Henson] | |
96 | ||
97 | *) Modify the way the V3 extension code looks up extensions. This now | |
98 | works in a similar way to the object code: we have some "standard" | |
99 | extensions in a static table which is searched with OBJ_bsearch() | |
100 | and the application can add dynamic ones if needed. The file | |
101 | crypto/x509v3/ext_dat.h now has the info: this file needs to be | |
102 | updated whenever a new extension is added to the core code and kept | |
103 | in ext_nid order. There is a simple program 'tabtest.c' which checks | |
104 | this. New extensions are not added too often so this file can readily | |
105 | be maintained manually. | |
106 | ||
107 | There are two big advantages in doing things this way. The extensions | |
108 | can be looked up immediately and no longer need to be "added" using | |
109 | X509V3_add_standard_extensions(): this function now does nothing. | |
110 | [Side note: I get *lots* of email saying the extension code doesn't | |
111 | work because people forget to call this function] | |
112 | Also no dynamic allocation is done unless new extensions are added: | |
113 | so if we don't add custom extensions there is no need to call | |
114 | X509V3_EXT_cleanup(). | |
115 | [Steve Henson] | |
116 | ||
117 | *) Modify enc utility's salting as follows: make salting the default. Add a | |
118 | magic header, so unsalted files fail gracefully instead of just decrypting | |
119 | to garbage. This is because not salting is a big security hole, so people | |
120 | should be discouraged from doing it. | |
121 | [Ben Laurie] | |
122 | ||
123 | *) Fixes and enhancements to the 'x509' utility. It allowed a message | |
124 | digest to be passed on the command line but it only used this | |
125 | parameter when signing a certificate. Modified so all relevant | |
126 | operations are affected by the digest parameter including the | |
127 | -fingerprint and -x509toreq options. Also -x509toreq choked if a | |
128 | DSA key was used because it didn't fix the digest. | |
129 | [Steve Henson] | |
130 | ||
131 | *) Initial certificate chain verify code. Currently tests the untrusted | |
132 | certificates for consistency with the verify purpose (which is set | |
133 | when the X509_STORE_CTX structure is set up) and checks the pathlength. | |
134 | ||
135 | There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour: | |
136 | this is because it will reject chains with invalid extensions whereas | |
137 | every previous version of OpenSSL and SSLeay made no checks at all. | |
138 | ||
139 | Trust code: checks the root CA for the relevant trust settings. Trust | |
140 | settings have an initial value consistent with the verify purpose: e.g. | |
141 | if the verify purpose is for SSL client use it expects the CA to be | |
142 | trusted for SSL client use. However the default value can be changed to | |
143 | permit custom trust settings: one example of this would be to only trust | |
144 | certificates from a specific "secure" set of CAs. | |
145 | ||
146 | Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions | |
147 | which should be used for version portability: especially since the | |
148 | verify structure is likely to change more often now. | |
149 | ||
150 | SSL integration. Add purpose and trust to SSL_CTX and SSL and functions | |
151 | to set them. If not set then assume SSL clients will verify SSL servers | |
152 | and vice versa. | |
153 | ||
154 | Two new options to the verify program: -untrusted allows a set of | |
155 | untrusted certificates to be passed in and -purpose which sets the | |
156 | intended purpose of the certificate. If a purpose is set then the | |
157 | new chain verify code is used to check extension consistency. | |
158 | [Steve Henson] | |
159 | ||
160 | *) Support for the authority information access extension. | |
161 | [Steve Henson] | |
162 | ||
163 | *) Modify RSA and DSA PEM read routines to transparently handle | |
164 | PKCS#8 format private keys. New *_PUBKEY_* functions that handle | |
165 | public keys in a format compatible with certificate | |
166 | SubjectPublicKeyInfo structures. Unfortunately there were already | |
167 | functions called *_PublicKey_* which used various odd formats so | |
168 | these are retained for compatibility: however the DSA variants were | |
169 | never in a public release so they have been deleted. Changed dsa/rsa | |
170 | utilities to handle the new format: note no releases ever handled public | |
171 | keys so we should be OK. | |
172 | ||
173 | The primary motivation for this change is to avoid the same fiasco | |
174 | that dogs private keys: there are several incompatible private key | |
175 | formats some of which are standard and some OpenSSL specific and | |
176 | require various evil hacks to allow partial transparent handling and | |
177 | even then it doesn't work with DER formats. Given the option anything | |
178 | other than PKCS#8 should be dumped: but the other formats have to | |
179 | stay in the name of compatibility. | |
180 | ||
181 | With public keys and the benefit of hindsight one standard format | |
182 | is used which works with EVP_PKEY, RSA or DSA structures: though | |
183 | it clearly returns an error if you try to read the wrong kind of key. | |
184 | ||
185 | Added a -pubkey option to the 'x509' utility to output the public key. | |
186 | Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*() and add | |
187 | EVP_PKEY_rset_*() functions that do the same as the EVP_PKEY_assign_*() | |
188 | except they up the reference count of the added key (they don't "swallow" | |
189 | the supplied key). | |
190 | [Steve Henson] | |
191 | ||
192 | *) Fixes to crypto/x509/by_file.c the code to read in certificates and | |
193 | CRLs would fail if the file contained no certificates or no CRLs: | |
194 | added a new function to read in both types and return the number | |
195 | read: this means that if none are read it will be an error. The | |
196 | DER versions of the certificate and CRL reader would always fail | |
197 | because it isn't possible to mix certificates and CRLs in DER format | |
198 | without choking one or the other routine. Changed this to just read | |
199 | a certificate: this is the best we can do. Also modified the code | |
200 | in apps/verify.c to take notice of return codes: it was previously | |
201 | attempting to read in certificates from NULL pointers and ignoring | |
202 | any errors: this is one reason why the cert and CRL reader seemed | |
203 | to work. It doesn't check return codes from the default certificate | |
204 | routines: these may well fail if the certificates aren't installed. | |
205 | [Steve Henson] | |
206 | ||
207 | *) Code to support otherName option in GeneralName. | |
208 | [Steve Henson] | |
209 | ||
210 | *) First update to verify code. Change the verify utility | |
211 | so it warns if it is passed a self signed certificate: | |
212 | for consistency with the normal behaviour. X509_verify | |
213 | has been modified to it will now verify a self signed | |
214 | certificate if *exactly* the same certificate appears | |
215 | in the store: it was previously impossible to trust a | |
216 | single self signed certificate. This means that: | |
217 | openssl verify ss.pem | |
218 | now gives a warning about a self signed certificate but | |
219 | openssl verify -CAfile ss.pem ss.pem | |
220 | is OK. | |
221 | [Steve Henson] | |
222 | ||
223 | *) For servers, store verify_result in SSL_SESSION data structure | |
224 | (and add it to external session representation). | |
225 | This is needed when client certificate verifications fails, | |
226 | but an application-provided verification callback (set by | |
227 | SSL_CTX_set_cert_verify_callback) allows accepting the session | |
228 | anyway (i.e. leaves x509_store_ctx->error != X509_V_OK | |
229 | but returns 1): When the session is reused, we have to set | |
230 | ssl->verify_result to the appropriate error code to avoid | |
231 | security holes. | |
232 | [Bodo Moeller, problem pointed out by Lutz Jaenicke] | |
233 | ||
234 | *) Fix a bug in the new PKCS#7 code: it didn't consider the | |
235 | case in PKCS7_dataInit() where the signed PKCS7 structure | |
236 | didn't contain any existing data because it was being created. | |
237 | [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson] | |
238 | ||
239 | *) Add a salt to the key derivation routines in enc.c. This | |
240 | forms the first 8 bytes of the encrypted file. Also add a | |
241 | -S option to allow a salt to be input on the command line. | |
242 | [Steve Henson] | |
243 | ||
244 | *) New function X509_cmp(). Oddly enough there wasn't a function | |
245 | to compare two certificates. We do this by working out the SHA1 | |
246 | hash and comparing that. X509_cmp() will be needed by the trust | |
247 | code. | |
248 | [Steve Henson] | |
249 | ||
250 | *) Correctly increment the reference count in the SSL_SESSION pointer | |
251 | returned from SSL_get_session(). | |
252 | [Geoff Thorpe <geoff@eu.c2.net>] | |
253 | ||
254 | *) Fix for 'req': it was adding a null to request attributes. | |
255 | Also change the X509_LOOKUP and X509_INFO code to handle | |
256 | certificate auxiliary information. | |
257 | [Steve Henson] | |
258 | ||
259 | *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document | |
260 | the 'enc' command. | |
261 | [Steve Henson] | |
262 | ||
263 | *) Add the possibility to add extra information to the memory leak | |
264 | detecting output, to form tracebacks, showing from where each | |
265 | allocation was originated. Also updated sid code to be multi- | |
266 | thread-safe. | |
267 | [Richard Levitte] | |
268 | ||
269 | *) Add options -text and -noout to pkcs7 utility and delete the | |
270 | encryption options which never did anything. Update docs. | |
271 | [Steve Henson] | |
272 | ||
273 | *) Add options to some of the utilities to allow the pass phrase | |
274 | to be included on either the command line (not recommended on | |
275 | OSes like Unix) or read from the environment. Update the | |
276 | manpages and fix a few bugs. | |
277 | [Steve Henson] | |
278 | ||
279 | *) Add a few manpages for some of the openssl commands. | |
280 | [Steve Henson] | |
281 | ||
282 | *) Fix the -revoke option in ca. It was freeing up memory twice, | |
283 | leaking and not finding already revoked certificates. | |
284 | [Steve Henson] | |
285 | ||
286 | *) Extensive changes to support certificate auxiliary information. | |
287 | This involves the use of X509_CERT_AUX structure and X509_AUX | |
288 | functions. An X509_AUX function such as PEM_read_X509_AUX() | |
289 | can still read in a certificate file in the usual way but it | |
290 | will also read in any additional "auxiliary information". By | |
291 | doing things this way a fair degree of compatibility can be | |
292 | retained: existing certificates can have this information added | |
293 | using the new 'x509' options. | |
294 | ||
295 | Current auxiliary information includes an "alias" and some trust | |
296 | settings. The trust settings will ultimately be used in enhanced | |
297 | certificate chain verification routines: currently a certificate | |
298 | can only be trusted if it is self signed and then it is trusted | |
299 | for all purposes. | |
300 | [Steve Henson] | |
301 | ||
302 | *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD). The | |
303 | problem was that one of the replacement routines had not been working since | |
304 | SSLeay releases. For now the offending routine has been replaced with | |
305 | non-optimised assembler. Even so, this now gives around 95% performance | |
306 | improvement for 1024 bit RSA signs. | |
307 | [Mark Cox] | |
308 | ||
309 | *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2 | |
310 | handling. Most clients have the effective key size in bits equal to | |
311 | the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key. | |
312 | A few however don't do this and instead use the size of the decrypted key | |
313 | to determine the RC2 key length and the AlgorithmIdentifier to determine | |
314 | the effective key length. In this case the effective key lenth can still | |
315 | be 40 bits but the key length can be 168 bits for example. This is fixed | |
316 | by manually forcing an RC2 key into the EVP_PKEY structure because the | |
317 | EVP code can't currently handle unusual RC2 key sizes: it always assumes | |
318 | the key length and effective key length are equal. | |
319 | [Steve Henson] | |
320 | ||
321 | *) Add a bunch of functions that should simplify the creation of | |
322 | X509_NAME structures. Now you should be able to do: | |
323 | X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0); | |
324 | and have it automatically work out the correct field type and fill in | |
325 | the structures. The more adventurous can try: | |
326 | X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0); | |
327 | and it will (hopefully) work out the correct multibyte encoding. | |
328 | [Steve Henson] | |
329 | ||
330 | *) Change the 'req' utility to use the new field handling and multibyte | |
331 | copy routines. Before the DN field creation was handled in an ad hoc | |
332 | way in req, ca, and x509 which was rather broken and didn't support | |
333 | BMPStrings or UTF8Strings. Since some software doesn't implement | |
334 | BMPStrings or UTF8Strings yet, they can be enabled using the config file | |
335 | using the dirstring_type option. See the new comment in the default | |
336 | openssl.cnf for more info. | |
337 | [Steve Henson] | |
338 | ||
339 | *) Make crypto/rand/md_rand.c more robust: | |
340 | - Assure unique random numbers after fork(). | |
341 | - Make sure that concurrent threads access the global counter and | |
342 | md serializably so that we never lose entropy in them | |
343 | or use exactly the same state in multiple threads. | |
344 | Access to the large state is not always serializable because | |
345 | the additional locking could be a performance killer, and | |
346 | md should be large enough anyway. | |
347 | [Bodo Moeller] | |
348 | ||
349 | *) New file apps/app_rand.c with commonly needed functionality | |
350 | for handling the random seed file. | |
351 | ||
352 | Use the random seed file in some applications that previously did not: | |
353 | ca, | |
354 | dsaparam -genkey (which also ignored its '-rand' option), | |
355 | s_client, | |
356 | s_server, | |
357 | x509 (when signing). | |
358 | Except on systems with /dev/urandom, it is crucial to have a random | |
359 | seed file at least for key creation, DSA signing, and for DH exchanges; | |
360 | for RSA signatures we could do without one. | |
361 | ||
362 | gendh and gendsa (unlike genrsa) used to read only the first byte | |
363 | of each file listed in the '-rand' option. The function as previously | |
364 | found in genrsa is now in app_rand.c and is used by all programs | |
365 | that support '-rand'. | |
366 | [Bodo Moeller] | |
367 | ||
368 | *) In RAND_write_file, use mode 0600 for creating files; | |
369 | don't just chmod when it may be too late. | |
370 | [Bodo Moeller] | |
371 | ||
372 | *) Report an error from X509_STORE_load_locations | |
373 | when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed. | |
374 | [Bill Perry] | |
375 | ||
376 | *) New function ASN1_mbstring_copy() this copies a string in either | |
377 | ASCII, Unicode, Universal (4 bytes per character) or UTF8 format | |
378 | into an ASN1_STRING type. A mask of permissible types is passed | |
379 | and it chooses the "minimal" type to use or an error if not type | |
380 | is suitable. | |
381 | [Steve Henson] | |
382 | ||
383 | *) Add function equivalents to the various macros in asn1.h. The old | |
384 | macros are retained with an M_ prefix. Code inside the library can | |
385 | use the M_ macros. External code (including the openssl utility) | |
386 | should *NOT* in order to be "shared library friendly". | |
387 | [Steve Henson] | |
388 | ||
389 | *) Add various functions that can check a certificate's extensions | |
390 | to see if it usable for various purposes such as SSL client, | |
391 | server or S/MIME and CAs of these types. This is currently | |
392 | VERY EXPERIMENTAL but will ultimately be used for certificate chain | |
393 | verification. Also added a -purpose flag to x509 utility to | |
394 | print out all the purposes. | |
395 | [Steve Henson] | |
396 | ||
397 | *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated | |
398 | functions. | |
399 | [Steve Henson] | |
400 | ||
401 | *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search | |
402 | for, obtain and decode and extension and obtain its critical flag. | |
403 | This allows all the necessary extension code to be handled in a | |
404 | single function call. | |
405 | [Steve Henson] | |
406 | ||
407 | *) RC4 tune-up featuring 30-40% performance improvement on most RISC | |
408 | platforms. See crypto/rc4/rc4_enc.c for further details. | |
409 | [Andy Polyakov] | |
410 | ||
411 | *) New -noout option to asn1parse. This causes no output to be produced | |
412 | its main use is when combined with -strparse and -out to extract data | |
413 | from a file (which may not be in ASN.1 format). | |
414 | [Steve Henson] | |
415 | ||
416 | *) Fix for pkcs12 program. It was hashing an invalid certificate pointer | |
417 | when producing the local key id. | |
418 | [Richard Levitte <levitte@stacken.kth.se>] | |
419 | ||
420 | *) New option -dhparam in s_server. This allows a DH parameter file to be | |
421 | stated explicitly. If it is not stated then it tries the first server | |
422 | certificate file. The previous behaviour hard coded the filename | |
423 | "server.pem". | |
424 | [Steve Henson] | |
425 | ||
426 | *) Add -pubin and -pubout options to the rsa and dsa commands. These allow | |
427 | a public key to be input or output. For example: | |
428 | openssl rsa -in key.pem -pubout -out pubkey.pem | |
429 | Also added necessary DSA public key functions to handle this. | |
430 | [Steve Henson] | |
431 | ||
432 | *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained | |
433 | in the message. This was handled by allowing | |
434 | X509_find_by_issuer_and_serial() to tolerate a NULL passed to it. | |
435 | [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>] | |
436 | ||
437 | *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null | |
438 | to the end of the strings whereas this didn't. This would cause problems | |
439 | if strings read with d2i_ASN1_bytes() were later modified. | |
440 | [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>] | |
441 | ||
442 | *) Fix for base64 decode bug. When a base64 bio reads only one line of | |
443 | data and it contains EOF it will end up returning an error. This is | |
444 | caused by input 46 bytes long. The cause is due to the way base64 | |
445 | BIOs find the start of base64 encoded data. They do this by trying a | |
446 | trial decode on each line until they find one that works. When they | |
447 | do a flag is set and it starts again knowing it can pass all the | |
448 | data directly through the decoder. Unfortunately it doesn't reset | |
449 | the context it uses. This means that if EOF is reached an attempt | |
450 | is made to pass two EOFs through the context and this causes the | |
451 | resulting error. This can also cause other problems as well. As is | |
452 | usual with these problems it takes *ages* to find and the fix is | |
453 | trivial: move one line. | |
454 | [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ] | |
455 | ||
456 | *) Ugly workaround to get s_client and s_server working under Windows. The | |
457 | old code wouldn't work because it needed to select() on sockets and the | |
458 | tty (for keypresses and to see if data could be written). Win32 only | |
459 | supports select() on sockets so we select() with a 1s timeout on the | |
460 | sockets and then see if any characters are waiting to be read, if none | |
461 | are present then we retry, we also assume we can always write data to | |
462 | the tty. This isn't nice because the code then blocks until we've | |
463 | received a complete line of data and it is effectively polling the | |
464 | keyboard at 1s intervals: however it's quite a bit better than not | |
465 | working at all :-) A dedicated Windows application might handle this | |
466 | with an event loop for example. | |
467 | [Steve Henson] | |
468 | ||
469 | *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign | |
470 | and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions | |
471 | will be called when RSA_sign() and RSA_verify() are used. This is useful | |
472 | if rsa_pub_dec() and rsa_priv_enc() equivalents are not available. | |
473 | For this to work properly RSA_public_decrypt() and RSA_private_encrypt() | |
474 | should *not* be used: RSA_sign() and RSA_verify() must be used instead. | |
475 | This necessitated the support of an extra signature type NID_md5_sha1 | |
476 | for SSL signatures and modifications to the SSL library to use it instead | |
477 | of calling RSA_public_decrypt() and RSA_private_encrypt(). | |
478 | [Steve Henson] | |
479 | ||
480 | *) Add new -verify -CAfile and -CApath options to the crl program, these | |
481 | will lookup a CRL issuers certificate and verify the signature in a | |
482 | similar way to the verify program. Tidy up the crl program so it | |
483 | no longer acesses structures directly. Make the ASN1 CRL parsing a bit | |
484 | less strict. It will now permit CRL extensions even if it is not | |
485 | a V2 CRL: this will allow it to tolerate some broken CRLs. | |
486 | [Steve Henson] | |
487 | ||
488 | *) Initialize all non-automatic variables each time one of the openssl | |
489 | sub-programs is started (this is necessary as they may be started | |
490 | multiple times from the "OpenSSL>" prompt). | |
491 | [Lennart Bang, Bodo Moeller] | |
492 | ||
493 | *) Preliminary compilation option RSA_NULL which disables RSA crypto without | |
494 | removing all other RSA functionality (this is what NO_RSA does). This | |
495 | is so (for example) those in the US can disable those operations covered | |
496 | by the RSA patent while allowing storage and parsing of RSA keys and RSA | |
497 | key generation. | |
498 | [Steve Henson] | |
499 | ||
500 | *) Non-copying interface to BIO pairs. | |
501 | (still largely untested) | |
502 | [Bodo Moeller] | |
503 | ||
504 | *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive | |
505 | ASCII string. This was handled independently in various places before. | |
506 | [Steve Henson] | |
507 | ||
508 | *) New functions UTF8_getc() and UTF8_putc() that parse and generate | |
509 | UTF8 strings a character at a time. | |
510 | [Steve Henson] | |
511 | ||
512 | *) Use client_version from client hello to select the protocol | |
513 | (s23_srvr.c) and for RSA client key exchange verification | |
514 | (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications. | |
515 | [Bodo Moeller] | |
516 | ||
517 | *) Add various utility functions to handle SPKACs, these were previously | |
518 | handled by poking round in the structure internals. Added new function | |
519 | NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to | |
520 | print, verify and generate SPKACs. Based on an original idea from | |
521 | Massimiliano Pala <madwolf@comune.modena.it> but extensively modified. | |
522 | [Steve Henson] | |
523 | ||
524 | *) RIPEMD160 is operational on all platforms and is back in 'make test'. | |
525 | [Andy Polyakov] | |
526 | ||
527 | *) Allow the config file extension section to be overwritten on the | |
528 | command line. Based on an original idea from Massimiliano Pala | |
529 | <madwolf@comune.modena.it>. The new option is called -extensions | |
530 | and can be applied to ca, req and x509. Also -reqexts to override | |
531 | the request extensions in req and -crlexts to override the crl extensions | |
532 | in ca. | |
533 | [Steve Henson] | |
534 | ||
535 | *) Add new feature to the SPKAC handling in ca. Now you can include | |
536 | the same field multiple times by preceding it by "XXXX." for example: | |
537 | 1.OU="Unit name 1" | |
538 | 2.OU="Unit name 2" | |
539 | this is the same syntax as used in the req config file. | |
540 | [Steve Henson] | |
541 | ||
542 | *) Allow certificate extensions to be added to certificate requests. These | |
543 | are specified in a 'req_extensions' option of the req section of the | |
544 | config file. They can be printed out with the -text option to req but | |
545 | are otherwise ignored at present. | |
546 | [Steve Henson] | |
547 | ||
548 | *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first | |
549 | data read consists of only the final block it would not decrypted because | |
550 | EVP_CipherUpdate() would correctly report zero bytes had been decrypted. | |
551 | A misplaced 'break' also meant the decrypted final block might not be | |
552 | copied until the next read. | |
553 | [Steve Henson] | |
554 | ||
555 | *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added | |
556 | a few extra parameters to the DH structure: these will be useful if | |
557 | for example we want the value of 'q' or implement X9.42 DH. | |
558 | [Steve Henson] | |
559 | ||
560 | *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and | |
561 | provides hooks that allow the default DSA functions or functions on a | |
562 | "per key" basis to be replaced. This allows hardware acceleration and | |
563 | hardware key storage to be handled without major modification to the | |
564 | library. Also added low level modexp hooks and CRYPTO_EX structure and | |
565 | associated functions. | |
566 | [Steve Henson] | |
567 | ||
568 | *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO | |
569 | as "read only": it can't be written to and the buffer it points to will | |
570 | not be freed. Reading from a read only BIO is much more efficient than | |
571 | a normal memory BIO. This was added because there are several times when | |
572 | an area of memory needs to be read from a BIO. The previous method was | |
573 | to create a memory BIO and write the data to it, this results in two | |
574 | copies of the data and an O(n^2) reading algorithm. There is a new | |
575 | function BIO_new_mem_buf() which creates a read only memory BIO from | |
576 | an area of memory. Also modified the PKCS#7 routines to use read only | |
577 | memory BIOSs. | |
578 | [Steve Henson] | |
579 | ||
580 | *) Bugfix: ssl23_get_client_hello did not work properly when called in | |
581 | state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of | |
582 | a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read, | |
583 | but a retry condition occured while trying to read the rest. | |
584 | [Bodo Moeller] | |
585 | ||
586 | *) The PKCS7_ENC_CONTENT_new() function was setting the content type as | |
587 | NID_pkcs7_encrypted by default: this was wrong since this should almost | |
588 | always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle | |
589 | the encrypted data type: this is a more sensible place to put it and it | |
590 | allows the PKCS#12 code to be tidied up that duplicated this | |
591 | functionality. | |
592 | [Steve Henson] | |
593 | ||
594 | *) Changed obj_dat.pl script so it takes its input and output files on | |
595 | the command line. This should avoid shell escape redirection problems | |
596 | under Win32. | |
597 | [Steve Henson] | |
598 | ||
599 | *) Initial support for certificate extension requests, these are included | |
600 | in things like Xenroll certificate requests. Included functions to allow | |
601 | extensions to be obtained and added. | |
602 | [Steve Henson] | |
603 | ||
604 | *) -crlf option to s_client and s_server for sending newlines as | |
605 | CRLF (as required by many protocols). | |
606 | [Bodo Moeller] | |
607 | ||
608 | Changes between 0.9.3a and 0.9.4 [09 Aug 1999] | |
609 | ||
610 | *) Install libRSAglue.a when OpenSSL is built with RSAref. | |
611 | [Ralf S. Engelschall] | |
612 | ||
613 | *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency. | |
614 | [Andrija Antonijevic <TheAntony2@bigfoot.com>] | |
615 | ||
616 | *) Fix -startdate and -enddate (which was missing) arguments to 'ca' | |
617 | program. | |
618 | [Steve Henson] | |
619 | ||
620 | *) New function DSA_dup_DH, which duplicates DSA parameters/keys as | |
621 | DH parameters/keys (q is lost during that conversion, but the resulting | |
622 | DH parameters contain its length). | |
623 | ||
624 | For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is | |
625 | much faster than DH_generate_parameters (which creates parameters | |
626 | where p = 2*q + 1), and also the smaller q makes DH computations | |
627 | much more efficient (160-bit exponentiation instead of 1024-bit | |
628 | exponentiation); so this provides a convenient way to support DHE | |
629 | ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of | |
630 | utter importance to use | |
631 | SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE); | |
632 | or | |
633 | SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE); | |
634 | when such DH parameters are used, because otherwise small subgroup | |
635 | attacks may become possible! | |
636 | [Bodo Moeller] | |
637 | ||
638 | *) Avoid memory leak in i2d_DHparams. | |
639 | [Bodo Moeller] | |
640 | ||
641 | *) Allow the -k option to be used more than once in the enc program: | |
642 | this allows the same encrypted message to be read by multiple recipients. | |
643 | [Steve Henson] | |
644 | ||
645 | *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts | |
646 | an ASN1_OBJECT to a text string. If the "no_name" parameter is set then | |
647 | it will always use the numerical form of the OID, even if it has a short | |
648 | or long name. | |
649 | [Steve Henson] | |
650 | ||
651 | *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp | |
652 | method only got called if p,q,dmp1,dmq1,iqmp components were present, | |
653 | otherwise bn_mod_exp was called. In the case of hardware keys for example | |
654 | no private key components need be present and it might store extra data | |
655 | in the RSA structure, which cannot be accessed from bn_mod_exp. | |
656 | By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for | |
657 | private key operations. | |
658 | [Steve Henson] | |
659 | ||
660 | *) Added support for SPARC Linux. | |
661 | [Andy Polyakov] | |
662 | ||
663 | *) pem_password_cb function type incompatibly changed from | |
664 | typedef int pem_password_cb(char *buf, int size, int rwflag); | |
665 | to | |
666 | ....(char *buf, int size, int rwflag, void *userdata); | |
667 | so that applications can pass data to their callbacks: | |
668 | The PEM[_ASN1]_{read,write}... functions and macros now take an | |
669 | additional void * argument, which is just handed through whenever | |
670 | the password callback is called. | |
671 | [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller] | |
672 | ||
673 | New function SSL_CTX_set_default_passwd_cb_userdata. | |
674 | ||
675 | Compatibility note: As many C implementations push function arguments | |
676 | onto the stack in reverse order, the new library version is likely to | |
677 | interoperate with programs that have been compiled with the old | |
678 | pem_password_cb definition (PEM_whatever takes some data that | |
679 | happens to be on the stack as its last argument, and the callback | |
680 | just ignores this garbage); but there is no guarantee whatsoever that | |
681 | this will work. | |
682 | ||
683 | *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=... | |
684 | (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused | |
685 | problems not only on Windows, but also on some Unix platforms. | |
686 | To avoid problematic command lines, these definitions are now in an | |
687 | auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl | |
688 | for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds). | |
689 | [Bodo Moeller] | |
690 | ||
691 | *) MIPS III/IV assembler module is reimplemented. | |
692 | [Andy Polyakov] | |
693 | ||
694 | *) More DES library cleanups: remove references to srand/rand and | |
695 | delete an unused file. | |
696 |