2 years agoFixup for 0053326492c1919 Added pbkdf password hash convenience functions... (forgot... master
Vladimir Sedach [Mon, 9 May 2011 13:45:29 +0000] 
Fixup for 0053326492c1919 Added pbkdf password hash convenience functions... (forgot to include a file in that patch).

2 years agoadd tree-hash test vector file to the test system
Nathan Froyd [Mon, 9 May 2011 02:01:31 +0000] 
add tree-hash test vector file to the test system

2 years agoAdded pbkdf password hash convenience functions, documentation, re-enabled pkcs5...
Vladimir Sedach [Tue, 3 May 2011 01:31:52 +0000] 
Added pbkdf password hash convenience functions, documentation, re-enabled pkcs5 unit tests.

2 years agoAdded function hex-string-to-byte-array
Vladimir Sedach [Tue, 3 May 2011 00:50:27 +0000] 
Added function hex-string-to-byte-array

2 years agoRemoved erroneous docstring from ironclad-error.
Vladimir Sedach [Tue, 3 May 2011 00:39:33 +0000] 
Removed erroneous docstring from ironclad-error.

3 years agomake {ENCRYPT,DECRYPT}-with-mode not take &KEY
Nathan Froyd [Thu, 3 Feb 2011 20:45:55 +0000] 
make {ENCRYPT,DECRYPT}-with-mode not take &KEY

There's no reason for these sorts of internal functions to take &KEY
arguments.  Make the functions fixed-airty and remove a bit of code
duplication in the process.

3 years agofix issues with SBCL, 32-bit platforms, and SHA512
Nathan Froyd [Mon, 29 Nov 2010 18:32:53 +0000] 
fix issues with SBCL, 32-bit platforms, and SHA512

We were attempting to bit-bash (SIMPLE-ARRAY (UNSIGNED-BYTE 64) (*)) on
32-bit platforms, and that doesn't work so well.  Rework the test for
using bit-bashing to avoid this situation.
# Please enter the commit message for your changes. Lines starting
# with '#' will be ignored, and an empty message aborts the commit.
# On branch mdx-digest-superclass
# Changes to be committed:
#   (use "git reset HEAD <file>..." to unstage)
# modified:   src/digests/digest.lisp

3 years agobreak out core of (UPDATE-DIGEST (X VECTOR)) to a common function
Nathan Froyd [Sat, 11 Sep 2010 19:44:06 +0000] 
break out core of (UPDATE-DIGEST (X VECTOR)) to a common function

This change requires making the code more generic at a slight runtime
performance cost.  I think the assembly changes will at least compensate
for this change on appropriate implementations.  For those
implementations where said assembly changes do not exist, patches
welcome.  (Those implementations may not have cared terribly much about
the performance of Ironclad, anyway.)

3 years agoadd x86-64 assembly for SHA1's expand function
Nathan Froyd [Sat, 11 Sep 2010 19:03:44 +0000] 
add x86-64 assembly for SHA1's expand function

3 years agoeliminate extraneous SHA256/512 constructors with &AUX
Nathan Froyd [Sat, 11 Sep 2010 03:58:35 +0000] 
eliminate extraneous SHA256/512 constructors with &AUX

3 years agomake all %MAKE-FOO-DIGEST constructors zero-arg functions
Nathan Froyd [Sat, 11 Sep 2010 03:55:53 +0000] 
make all %MAKE-FOO-DIGEST constructors zero-arg functions

3 years agopush toplevel MACROLETs down for digests and des
Nathan Froyd [Sat, 11 Sep 2010 03:47:26 +0000] 
push toplevel MACROLETs down for digests and des

3 years agoadd assembly version of SHA256's expand function
Nathan Froyd [Sat, 11 Sep 2010 03:08:51 +0000] 
add assembly version of SHA256's expand function

3 years agoupdate NEWS with bugfixes and improvements
Nathan Froyd [Sat, 11 Sep 2010 03:08:02 +0000] 
update NEWS with bugfixes and improvements

3 years agofix x86-64 sbcl bug with overzealous use of N-WORD-BYTES
Nathan Froyd [Sat, 11 Sep 2010 02:06:54 +0000] 
fix x86-64 sbcl bug with overzealous use of N-WORD-BYTES

3 years agoadd loop unrolling support for fill-ub8-* VOPs
Nathan Froyd [Fri, 10 Sep 2010 23:35:28 +0000] 
add loop unrolling support for fill-ub8-* VOPs

3 years agoadd x86oid optimizations for SHA1's EXPAND-BLOCK
Nathan Froyd [Sun, 25 Apr 2010 23:47:03 +0000] 
add x86oid optimizations for SHA1's EXPAND-BLOCK

3 years agoadd x86oid optimizations for FILL-BLOCK-* functions
Nathan Froyd [Sun, 25 Apr 2010 22:49:31 +0000] 
add x86oid optimizations for FILL-BLOCK-* functions

3 years agoadd SBCL-OPT module
Nathan Froyd [Sun, 25 Apr 2010 22:48:49 +0000] 
add SBCL-OPT module

For now, we'll optimize some of the digest functions.

3 years agofix ENCRYPTED-MESSAGE-LENGTH for ECB and CBC modes
Nathan Froyd [Fri, 18 Jun 2010 21:39:32 +0000] 

3 years agoremove unused TAGBODY tag in adler32
Nathan Froyd [Fri, 18 Jun 2010 21:37:32 +0000] 
remove unused TAGBODY tag in adler32

4 years agodelete unneeded DEFCLASS DIGEST
Nathan Froyd [Tue, 20 Apr 2010 19:10:27 +0000] 
delete unneeded DEFCLASS DIGEST

4 years agoadd 64-bit Tiger support for x86-64 SBCL
Nathan Froyd [Sat, 17 Apr 2010 04:41:29 +0000] 
add 64-bit Tiger support for x86-64 SBCL

4 years agoupdate everything to version 0.28
Nathan Froyd [Sat, 27 Mar 2010 01:19:55 +0000] 
update everything to version 0.28

4 years agoput ironclad-tests system back in ironclad.asd
Nathan Froyd [Sat, 27 Mar 2010 01:15:31 +0000] 
put ironclad-tests system back in ironclad.asd

4 years agoexport MAKE-TIGER-TREE-HASH
Nathan Froyd [Sat, 27 Mar 2010 01:13:11 +0000] 

4 years agodocument tree hashes in the manual
Nathan Froyd [Sat, 27 Mar 2010 01:09:27 +0000] 
document tree hashes in the manual

4 years agoupdate NEWS with size and tree hash improvements
Nathan Froyd [Sat, 27 Mar 2010 00:54:20 +0000] 
update NEWS with size and tree hash improvements

4 years agotweak platform information slightly
Nathan Froyd [Sat, 27 Mar 2010 00:50:00 +0000] 
tweak platform information slightly

4 years agoadd information about git repository
Nathan Froyd [Sat, 27 Mar 2010 00:49:09 +0000] 
add information about git repository

4 years agoTweak the TREE-HASH digest to froydnj's specifications:
Francois-Rene Rideau [Thu, 18 Mar 2010 20:10:09 +0000] 
Tweak the TREE-HASH digest to froydnj's specifications:
* Remove the :tth digest type.
* Just have MAKE-TIGER-TREE-HASH do the Right Thing.
* Remove the dependency on binascii in the test vectors.
* Instead, split tests between tree-hash.testvec and binascii tests.

4 years agoFix tests for TTH.
Francois-Rene Rideau [Fri, 5 Mar 2010 22:39:46 +0000] 
Fix tests for TTH.

4 years agoSeparate ironclad-tests, add test vectors for tth
Francois-Rene Rideau [Sun, 28 Feb 2010 03:59:27 +0000] 
Separate ironclad-tests, add test vectors for tth

4 years agoAdd support for (tiger-)tree-hash digests.
Francois-Rene Rideau [Sat, 27 Feb 2010 22:29:47 +0000] 
Add support for (tiger-)tree-hash digests.

4 years agomove XOR-BLOCK earlier in the compilation process
Nathan Froyd [Thu, 7 Jan 2010 04:18:38 +0000] 
move XOR-BLOCK earlier in the compilation process

HMAC depends on it, so it shouldn't be buried in the ciphers code.
Thanks to Jochen Schmidt for pointing this out.

4 years agodial back the inlining for non-mod-wordsize implementations
Nathan Froyd [Thu, 7 Jan 2010 04:14:43 +0000] 
dial back the inlining for non-mod-wordsize implementations

Inlining all the MOD32-FUNC primitives and functions that use them can
cause problems for implementations that don't support modular
arithmetic.  Plus there's no benefit in doing so; it's just code bloat.
Define appropriate features and liberally sprinkle conditionals in the
appropriate digest implementations.

4 years agoignore variables in MAKE-CIPHER compiler macro
Nathan Froyd [Thu, 7 Jan 2010 03:42:56 +0000] 
ignore variables in MAKE-CIPHER compiler macro

4 years agofix DECRYPT-WITH-MODE for padded cipher modes
Nathan Froyd [Tue, 24 Nov 2009 20:43:38 +0000] 
fix DECRYPT-WITH-MODE for padded cipher modes

Thanks to Stelian Ionescu for noticing this.

Nathan Froyd [Tue, 24 Nov 2009 20:41:55 +0000] 

Thanks to Stelian Ionescu for pointing this out.

4 years agomove MASSAGE-SYMBOL to a common place
Nathan Froyd [Sun, 15 Nov 2009 19:53:41 +0000] 
move MASSAGE-SYMBOL to a common place

Turns out ciphers--via (SETF %FIND-CIPHER)--and digests need it.  The
new directory structure hid that, and it must have worked by accident
earlier due to ASDF's dependency traversal.  Tidy up some dependencies
while we're at it.

4 years agofix little-endian INTEGER-TO-OCTETS
Nathan Froyd [Sun, 15 Nov 2009 19:45:35 +0000] 
fix little-endian INTEGER-TO-OCTETS

4 years agoupdate known-good platforms
Nathan Froyd [Sat, 14 Nov 2009 03:10:53 +0000] 
update known-good platforms

4 years agofix errors on 32-bit big-endian CMUCL and SBCL
Nathan Froyd [Sat, 14 Nov 2009 03:08:46 +0000] 
fix errors on 32-bit big-endian CMUCL and SBCL

Such platforms don't have specialized 64-bit arrays, so we were trying
to bit-bash SIMPLE-VECTORs, which doesn't work.  Thanks to Raymond Toy
for noticing this.

4 years agomake DEFINE-STREAM-{READ,WRITE}-SEQUENCE more general
Nathan Froyd [Sat, 14 Nov 2009 03:05:56 +0000] 

4 years agoInitial commit.
Nathan Froyd [Wed, 11 Nov 2009 14:03:15 +0000] 
Initial commit.