18 Commits

Author SHA1 Message Date
Geoff Thorpe
03ef63ba69 Try to catch some memory leaks in the error case. 2000-07-26 22:11:13 +00:00
Richard Levitte
a6f8bbcad9 Avoid the conflict between () and (void) 2000-07-12 15:14:12 +00:00
Richard Levitte
9188480827 Handle the hwcrhk private key with a bit more consistency. Also, for the
random numbers generator, always return status 1 since the entropy is
already presumably there...
2000-07-07 17:04:44 +00:00
Richard Levitte
64c4f5732d Add the possibility to load prvate and public keys from an engine and
implement it for nCipher hardware.  The interface in itself should be
clear enough, but the nCipher implementation is currently not the
best when it comes to getting a passphrase from the user.  However,
getting it better is a little hard until a better user interaction
method is create.

Also, use the possibility in req, so we can start to create CSR's with
keys from the nForce box.

WARNING: I've made *no* tests yet, mostly because I didn't implement
this on the machine where I have an nForce box to play with.  All I
know is that it compiles cleanly on Linux...
2000-07-06 18:40:10 +00:00
Richard Levitte
f3052a9eee Don't initialise the pointers to mutex functions directly in the structure.
This is correctly taken care of by hwcrhk_init().  While we're at it, give
this engine the official name of the library used (CHIL, for Cryptographic
Hardware Interface Library).
2000-07-05 16:00:18 +00:00
Richard Levitte
ae02fc5348 Make it possible to turn off compilation of hardware support through
the configuration parameter 'no-hw'.
2000-06-30 11:02:02 +00:00
Richard Levitte
3257904c56 It makes much more sense and is much more consistent with the rest of
OpenSSL to have to opt out hardware support instead of having to opt
it in.  And since the hardware support modules are self-contained and
actually check that the vendor stuff is loadable, it still works as
expected, or at least, so I think...
2000-06-29 21:20:14 +00:00
Richard Levitte
70d03c4f59 Make the use of logstream thread-safe. 2000-06-29 16:32:34 +00:00
Richard Levitte
d813a428a7 When closing the hwcrhk engine, also remove the reference to the
logstream.
2000-06-29 16:16:50 +00:00
Richard Levitte
3b2972d8d9 Add the possibility to control some engine internals. 2000-06-29 14:26:07 +00:00
Richard Levitte
2165d91196 Rename 'hwcrhk' to 'ncipher' in all public symbols. Redo the logging function so it takes a BIO. Make module-local functions static 2000-06-29 13:00:07 +00:00
Geoff Thorpe
665b5ab5a7 Strange how one line can make a world of difference, particularly when
the one line turns an error return value into a success return value.
:-) "openssl speed -engine hwcrhk rsa1024" now passes through ok.
2000-06-26 15:58:33 +00:00
Richard Levitte
b215f70a0e Merge of stuff from main trunk, all conflicts resolved, and addition
of dynamic lock support in the nCipher code.
2000-06-19 17:28:22 +00:00
Richard Levitte
7ed20a2158 Merge of stuff from main trunk, all conflicts resolved. 2000-06-19 14:44:57 +00:00
Geoff Thorpe
f18ef82a9f Little typo. 2000-06-14 17:54:28 +00:00
Richard Levitte
9a4051050c Geoff inspired me to nullify some pointers if initialisation went
wrong.  Additionally, just give a new value to hndidx once.
2000-06-14 16:57:57 +00:00
Geoff Thorpe
4c4ea428cc DSO_bind() is effectively a method-specific wrapper for dlopen() or
whatever the underlying API is. It must return (void *) because shared
libraries can expose functions, structures, or whatever. However, some
compilers give loads of warnings about casted function pointers through
this code, so I am explicitly casting them to the right prototypes.
2000-06-14 14:28:16 +00:00
Richard Levitte
86787f93d6 - merged in the latest from the main trunk, fixed all conflicts
- implemented nCipher support via the nfhwcrhk library (not well tested).
- make update + make depend
2000-06-13 16:21:06 +00:00