Commit Graph

414 Commits

Author SHA1 Message Date
ccremers
b56c01c422 - Added '--max-depth=X' switch (which is equal to the old '-l X -a')
- Modified semantics of -l with -a : this corresponds more to the
  intuition and introduces the new option to prune proofs based on trace
  length.
2004-12-09 15:11:45 +00:00
ccremers
4f36181c3c - Removed debugging stuff, now that the memory stuff is solved. It
turned out that I solved the memory leak first, and then spent an
  afternoon finding the 8 blocks. These were simply not being given back
  by memrealloc, which I should have guessed.
2004-12-09 13:34:36 +00:00
ccremers
c7d9517eac - Fixed some errors. No more memory leaks. 2004-12-09 13:23:26 +00:00
ccremers
c690b0622a - More cleanup and comments. 2004-12-08 19:30:26 +00:00
ccremers
1c5a9986f6 - Added many comments. 2004-12-08 16:41:43 +00:00
ccremers
3ca180d968 - Despite a full afternoon of debugging, semiRunCreate/Destroy still
lose 8 blocks. I'm fairly confused.
2004-12-08 16:25:27 +00:00
ccremers
f2a2c8ea14 - Moved role creation into the protocol creation. This will make it
easier to add MSC-style input to the input language compiler later.
2004-11-16 12:51:23 +00:00
ccremers
506e42f841 - Re-indented the files. 2004-11-16 12:07:55 +00:00
ccremers
a38925c9c2 - Added some useful macros to term.h to address subparts (e.g.
TermOp1(t)). Renamed all uses.
2004-11-16 12:06:36 +00:00
ccremers
343314896b - Added version info to compilation process. 2004-11-01 14:52:52 +00:00
ccremers
b6af2f9dac - First try for diff function. 2004-11-01 14:06:26 +00:00
ccremers
c195ab95a1 - Now in the wiki. 2004-10-30 10:07:21 +00:00
ccremers
bdc336d7f9 - Todo list is now handled by the wiki:
http://www.win.tue.nl/~ccremers/twiki/bin/view.pl/Scyther/ScytherBug
2004-10-29 14:15:35 +00:00
ccremers
aaa0d415f9 - Graph closure fixed. 2004-10-28 15:37:13 +00:00
ccremers
0ec70b9de0 - Added lots of debugging info. 2004-10-28 15:23:16 +00:00
ccremers
3673fc689d - Improved roledef printing by adding roledefPrintShort. 2004-10-28 12:56:13 +00:00
ccremers
61457b5f3d - Fixed the agentsOfRunPrint output. It was caused by the agent adding
order.
- Fixed the pruning bug, which was related to this.
2004-10-28 12:33:57 +00:00
ccremers
234edae741 - Added sort of bug report. 2004-10-27 16:32:44 +00:00
ccremers
9a24bb0f21 - Somehow, agentOfRunRole is playing up. I found a bug in the pruning
(untrusted actors), but the fix did not work. It's maybe due to the
  roleInstance variants.
2004-10-27 16:20:12 +00:00
ccremers
461a040d29 - Revised Arachne dot output significantly. It is now based on explicit
ranking instead of the subgraphs. This will make it easier to layout
  e.g. LaTeX MSCs using the same algorithm.
2004-10-27 16:10:58 +00:00
ccremers
2680a2ca7a - Added rank calculation and output. If the subgraphs are removed, this
will allow for better positioning of the graphs. It also helps a lot
  for latex output. In fact, latex output is fairly trivial now.
2004-10-25 14:28:53 +00:00
ccremers
3cda6e53fa - Another important issue. 2004-10-20 15:59:23 +00:00
ccremers
fce9fae9c3 - Removed warnings from the compiler for conflicting types: not to
stdout anymore, but now in the normal tradition of eprintf and
  globalError.
2004-10-19 12:03:40 +00:00
ccremers
7b3cb4dfb9 - To test. 2004-10-18 14:36:43 +00:00
ccremers
8a7369a84e - Some stuff about empty preceding label sets and Arachne. 2004-10-18 13:49:41 +00:00
ccremers
67673cb608 - No more warning for output in standard debug mode. 2004-10-18 13:49:20 +00:00
ccremers
19b3c74e65 - Remove obsolete child parameter. 2004-10-18 13:06:22 +00:00
ccremers
94b3ac7c96 - Added debug code for dot output.
- Push/pop goals are counted now, making the child parameter obsolete.
2004-10-18 13:04:34 +00:00
ccremers
4b10c7f151 - Added priorities. 2004-10-16 22:03:33 +00:00
ccremers
abad7044dd - Added thingy. 2004-10-15 19:53:04 +00:00
ccremers
795f28006d - Added some thoughts. 2004-10-14 20:49:49 +00:00
ccremers
d33ec486ce - Modified -l switch to also serve as proof depth limit. 2004-10-14 15:25:28 +00:00
ccremers
ba832159b1 - Added a new prioritylevel for seemingly public keys, but the splice-as
problem remains.
2004-10-14 15:09:48 +00:00
ccremers
9ac12f9198 - More serious problem found. 2004-10-14 14:33:22 +00:00
ccremers
123b12a1c0 - Problems with Splice/AS suggest heuristic improvements. 2004-10-14 14:30:27 +00:00
ccremers
87a75106d1 - Printing the protocol tab-separated from the role is more useful for
script-based parsing.
2004-10-14 13:29:28 +00:00
ccremers
2bc1df6135 - Improved readability of printed claims.
- Fixed comment.
2004-10-14 13:19:36 +00:00
ccremers
40584ba6bd - Noted some stuff about graph computations, so that I won't forget
this.
2004-10-12 18:08:01 +00:00
ccremers
487212a9f9 - The TMN protocol was wrongly reporting an error in the protocol. This
turned out to be caused by an over-protective first-read detection.
2004-10-12 16:58:29 +00:00
ccremers
d64badefdc - Found a problem with type flaw attacks. 2004-10-12 15:23:03 +00:00
ccremers
0de3320009 - Fixed a memory leak in termLocal. This did not cause any problems for
the modelchecker, as it calls it only once, but it caused major
  problems for the arachne engine, which creates and destroys semiruns
  all the time.
2004-10-12 15:12:20 +00:00
ccremers
3b4b367a4a - Minor correction. Probably redundant for a good compiler ;) 2004-09-20 17:41:53 +00:00
ccremers
02e99761ae - Split roleInstance into two more managable parts. 2004-09-20 12:40:01 +00:00
ccremers
be366afa0e - A good reduction idea for secrecy added to the todo list. 2004-09-07 09:56:06 +00:00
ccremers
8570465e48 - More todo. 2004-09-01 19:11:06 +00:00
ccremers
8b48aade68 - Huge effort to make match type 2 (typeflaw generic) matching work.
Problem with goals that turn into tuples, will have to be solved.
2004-08-31 14:31:06 +00:00
ccremers
0e9b7dcf11 - Some added error/bounds detection all around. 2004-08-31 12:35:05 +00:00
ccremers
a673ea4ad1 - Write down, so that we don't forget. 2004-08-31 08:58:50 +00:00
ccremers
f5ab30995c - Removed the debugging output. 2004-08-30 22:09:44 +00:00
ccremers
5c90522c55 - Fixed a bug in the pruning algorithm, where intruder runs were also
checked for agent lists, which is false.
2004-08-30 22:08:44 +00:00
ccremers
b04bc86185 - Some minor cleanups. 2004-08-30 21:49:51 +00:00
ccremers
4832e9116c - Added pruning theorem for untrusted actors. 2004-08-30 21:07:45 +00:00
ccremers
d43e3d432f - Ignoring singular variables seems to be a smart choice, although it
implies that the intruder can generate any type. That is not conform
  the usual semantics. So we either change the usual semantics (wise) or
  we make this choice optional.
2004-08-30 20:48:11 +00:00
ccremers
8f441ac913 - Fixed some minor issues.
- Fixed type flaw in labellist type.
2004-08-30 20:08:11 +00:00
ccremers
02041cfbab - Fixed binding displays.
- Improved attack dot output.
- goal_graph_create now takes originator assumption into account.
2004-08-30 13:57:16 +00:00
ccremers
1d431dc6f1 - Attack output is a bit broken now for Arachne. Fix. 2004-08-30 06:07:17 +00:00
ccremers
5035a35d51 - Bug spotted. 2004-08-28 17:28:14 +00:00
ccremers
25fa261e30 - Added some comments. 2004-08-28 14:05:38 +00:00
ccremers
c907c1f657 - Added prefixed start nodes to indicate agent initiative in dot output. 2004-08-28 14:00:48 +00:00
ccremers
08f2155527 - Denoting 'empty term' with '*' from now on, yields more compact
output.
2004-08-28 14:00:22 +00:00
ccremers
391c939b83 - New algorithm to draw bindings between runs. Much cleaner. 2004-08-28 13:47:37 +00:00
ccremers
b349b6cef2 - More improvements to the dot output. 2004-08-28 12:42:11 +00:00
ccremers
acb89922f1 - Singular variables need to be bound as well (to ensure ordering is
correct w.r.t. e.g. nonces, if the intruder cannot construct them.)
2004-08-28 12:20:50 +00:00
ccremers
2ddd1eee13 - Improved dot output for Arachne attacks. 2004-08-28 11:43:06 +00:00
ccremers
9d64b837db - Improved roledef printing for NULL, NULL roles (intruder)
- Added graph output in dot format.
2004-08-28 09:24:30 +00:00
ccremers
6c2730af1a - Added some todo stuff. 2004-08-27 19:29:41 +00:00
ccremers
4420e06e4e - Ignore choose actions when determining Arachne trace length. 2004-08-27 19:15:24 +00:00
ccremers
4f534410bd - Implemented ordering checks. Need some test to validate this though. 2004-08-27 19:06:15 +00:00
ccremers
957b920b98 - Added extra Arachne check for -r0. 2004-08-27 18:26:19 +00:00
ccremers
17ad6de97b - Semistate printing now reports trace length.
- Pruning was wrong, so the shortest attack wasn't always found. Now it
  is.
2004-08-27 18:18:16 +00:00
ccremers
198afa135e - Implemented attack length scanner per claim. Not stored yet. 2004-08-27 18:09:09 +00:00
ccremers
6ccb09297a - Better prune adherence. 2004-08-27 17:37:43 +00:00
ccremers
f90f16fe93 - Arachne engine now respects --prune=2 (and thus the default setting)
somewhat. There is no good definition of length yet, so we don't do
  this yet.
2004-08-27 17:35:23 +00:00
ccremers
21b2c27320 - Niagree claim seems to be working fine now. 2004-08-27 17:25:38 +00:00
ccremers
2decf44bd2 - Checks are now in. Untested though. 2004-08-27 15:02:33 +00:00
ccremers
68bbdc2794 - Added interfaces for the more interesting Arachne claim checks. 2004-08-27 14:48:58 +00:00
ccremers
fd3769d683 - Agreement test for Archne implemented (untested). 2004-08-27 14:41:06 +00:00
ccremers
4009ca86ed - Added some sanity checks for read/send/claim role parameters.
- The cl->roles are now distance-ordered. This, the first role is at
  distance 0, etc. This is useful for checking e.g. synchronisation.
2004-08-27 13:40:46 +00:00
ccremers
dfeaf83327 - Added 'termlistFind' function, which is more generic than inTermlist 2004-08-27 13:10:46 +00:00
ccremers
d8e0e93bcf - Fixed a condition check in termlistAddNew.
- Roles are now computed from prec for each claim.
2004-08-27 12:36:23 +00:00
ccremers
542044e36f - Added preliminary labellist support to the system. 2004-08-27 11:52:43 +00:00
ccremers
275743c1a3 - Fixed a bug where labels where not generated nicely if the symbols
already had been declared in another role.
2004-08-27 10:24:19 +00:00
ccremers
d58fc5ab43 - Made the label naming unique, by adding tuple info with the protocol
name. Now, we can simply test multiple protocol names by
  concatenation.
- Removed the pointer equality leaf hypothesis, as it didn't hold
  anymore.
2004-08-27 10:08:03 +00:00
ccremers
959c8d2c8b - Added termlist_to_tuple function. 2004-08-26 12:36:01 +00:00
ccremers
6c38253559 - Turned the exit codes into enum types, making it more generic. 2004-08-24 13:09:39 +00:00
ccremers
7d0be35658 - Bugfix: term output now correctly displays local constants of a run
before it is bound.
2004-08-23 13:46:48 +00:00
ccremers
0fc008fe33 - Added keylevels to symbols. This is to help pruning the proofs, for
terms and patterns that do not originate on regular nodes.
2004-08-20 19:16:56 +00:00
ccremers
98bff1e5e2 - Solved the TERM_Hidden issue for the keys in termMguSubTerm. Yields
cleaner behaviour for MguSubterm.
2004-08-20 15:09:49 +00:00
ccremers
d7e49028c1 - Added pruning of functions the intruder does not know (e.g. SK) 2004-08-20 14:55:34 +00:00
ccremers
851044ecd0 - Improved the SK lemma, but it is NOT correct yet. 2004-08-20 11:47:00 +00:00
ccremers
72d52a6e12 - Key goals now have priority. This strategy yields complete proofs for
e.g. bke, and reduces states for NSL.
2004-08-20 10:52:40 +00:00
ccremers
baae7ef94a - The proofs now also show a list of open goals at each step. 2004-08-20 09:53:44 +00:00
ccremers
bf2cbb5540 - Updated the todo list. 2004-08-20 09:26:34 +00:00
ccremers
be44ed047a - Fixed some goal selection issues.
- Added note about mirroring model checker semantics.
2004-08-20 09:21:39 +00:00
ccremers
7308791c83 - More todos. 2004-08-20 08:01:35 +00:00
ccremers
bd84625ae4 - Fixed some more problems. Seems to be stable, although pruning is not
sufficient. Investigate bke-broken.
2004-08-19 15:30:31 +00:00
ccremers
f2bc78cc1f - Improved proof output. 2004-08-19 14:55:21 +00:00
ccremers
8fa7c4e839 - Fixed bug in printing.
- Algorithm should work again.
2004-08-19 14:52:17 +00:00
ccremers
f25f0abd4e - Fixed a memory error. 2004-08-19 14:49:03 +00:00
ccremers
35c55c9483 - Fixed a bug for NULL case in interm/subterm.
- Fixed a bug where the mgu termlist was never deleted in
  interm/subterm.
2004-08-19 13:55:16 +00:00