Commit Graph

590 Commits

Author SHA1 Message Date
ccremers
00616e45ed - Bit masking was incorrect: & binds less strong than == !
This caused many of the --experimental switches not to work.
2006-02-28 13:41:36 +00:00
ccremers
cf832ca1b1 - Seems to work again, but further testing is needed. 2006-02-27 22:27:09 +00:00
ccremers
b49d13b6ee - [[[ Broken commit. ]]]
Stuff seems to be working again, slightly less efficient though (count
  states).
2006-02-27 16:08:17 +00:00
ccremers
bb16bd755e - Print states in a more countable format. 2006-02-27 15:20:37 +00:00
ccremers
c22173e5ee - [[[ Broken commit ]]]
More work on the arachne multiple-decryptor. Horrific.
2006-02-26 20:01:22 +00:00
ccremers
95df010a54 - [[[ Broken commit ]]]
More intermediate work.
2006-02-26 17:18:59 +00:00
ccremers
0ce88af6ac - [[[ Broken commit ]]]
Committing partial new Warshall work because it is getting too big.
2006-02-26 15:00:58 +00:00
ccremers
1d3d154a2f - If the timebound is hit, it should be reported anyway, because the
results are not to be trusted anymore.
2006-02-23 16:21:25 +00:00
ccremers
b16023bf0e - Cleaned up heuristic code. Note that there is a "hidden" heuristic:
implicitly, older goals are resolved first, if some goals have equal
  weights. This is encoded in the "w <=" comparison; if this is set to
  "w <", the heuristic becomes much less effective.
2006-02-23 15:03:43 +00:00
ccremers
f333fb8276 - Explicit casts. 2006-02-23 10:44:44 +00:00
ccremers
8f896432d1 - Added some stuff on the new graph code. 2006-02-22 16:58:11 +00:00
ccremers
f376260512 - Changed default heuristic to 34. 2006-02-22 15:57:55 +00:00
ccremers
41e797413c - Added new heuristics based on hidelevel results.
Preliminary results:
  1. For typed matching, either heuristic 32 or 34 are best, and far
  superior to the previous best (3).
  2. For untyped matching, partial tests indicate that heuristic 1 is
  best, which is rather interesting.
2006-02-22 15:48:58 +00:00
ccremers
921c82876d - experimental=4 now disables some things. Weirdly enough, they don't
seem to make much difference.
2006-02-22 09:53:50 +00:00
ccremers
dbc0a3583d - Better setup for --experimental= switch using bit masks. 2006-02-22 08:55:42 +00:00
ccremers
10b6793d97 - More cleaning of switches. 2006-02-22 08:47:22 +00:00
ccremers
5ddcdfed22 - Added --long-help and -E, --expert switches. 2006-02-22 08:41:06 +00:00
ccremers
b2e40e07f3 - Some more work on hidelevel backbone.
- Added '--count-states' switch for the Arachne engine.
2006-02-22 08:24:29 +00:00
ccremers
d3f2971181 - Reindent script was improved (and consequences added) 2006-02-21 21:35:14 +00:00
ccremers
5d2d836d07 - Much work for the skeleton of the Hidelevel lemma. 2006-02-21 20:29:05 +00:00
ccremers
bb7259a1ad - Removed some too interesting pruning methods that really need theorems
first. Revealed by the certified e-mail protocol by Abadi and
  Blanchet.
2006-01-17 16:18:26 +00:00
ccremers
3ed59b867a - Added an idea. 2006-01-17 12:30:16 +00:00
ccremers
f0715c030d - Stored good idea. 2006-01-09 11:56:44 +00:00
ccremers
baf1856943 - Added some tokens for future usage: 'function', 'hashfunction',
'knows', 'trusted'.
2006-01-09 09:38:17 +00:00
ccremers
8b30526a57 - Added a note about inversekeys in a role definition. 2006-01-07 13:28:13 +00:00
ccremers
96e7a32bff - Added '--unbounded' switch. 2006-01-06 12:46:04 +00:00
ccremers
99861d3e03 - Added '--untyped' switch which is to be preferred above the older
--match=2 notation.
2006-01-06 12:23:11 +00:00
ccremers
d9b0f412e4 - Added a good idea. 2006-01-06 12:11:32 +00:00
ccremers
92342683f5 - Cleanup of unused structure. 2006-01-03 11:34:48 +00:00
ccremers
c9eaf1f95f - Minor stuff. 2006-01-03 11:34:27 +00:00
ccremers
066bc810d8 - Notes on the warshall algorithm that is currently taking a third of
processing time, which is way too much.
2006-01-02 21:19:53 +00:00
ccremers
da75862d82 - Huge code documentation effort. 2006-01-02 21:06:08 +00:00
ccremers
6676266f4a - More refactoring to improve the code. 2006-01-02 20:18:47 +00:00
ccremers
e6505a72a3 - Further refactoring.
- Some cleanup.
2006-01-02 19:55:34 +00:00
ccremers
a5acc4984a - More refactoring for Arachne. Slowly we're getting somewhere. 2006-01-02 19:19:23 +00:00
ccremers
e592a0a432 - Refactoring code: splitting stuff out of arachne.c 2006-01-02 18:43:25 +00:00
ccremers
4023ef237e - Some reindentation. Nothing interesting, just syntax. 2006-01-02 16:07:56 +00:00
ccremers
e21627442a - Added 'singular' directive for roles. Syntax:
protocol ns3 (I,R)
  {
    singular role I:
    {
    }
  }
2006-01-02 16:05:53 +00:00
ccremers
724faa8949 - Cleaned up some printf's and warnings in compiler.c
- Added warning for unspecified roles.
2006-01-02 15:29:41 +00:00
ccremers
6516741983 - More notes. 2006-01-02 15:10:40 +00:00
ccremers
25fe5b210e - Idea about timestamps added. 2006-01-02 14:34:46 +00:00
ccremers
28f13aff26 - Added some final-day notes for 2005. 2005-12-31 19:34:50 +00:00
ccremers
441644e6d2 - Fixed the ECSS version copy. 2005-12-30 15:56:05 +00:00
ccremers
8c04a7517c - Added script to copy the latest Scyther version to the ecss
repository.
2005-12-30 15:54:32 +00:00
ccremers
d2ac518234 - Fixed typo in --help (noted by Sjouke)
- State-space should generate classes.
2005-12-30 15:11:39 +00:00
ccremers
6f670d7ab6 - In xml output, renamed 'attack' tag to 'state', which in general
reflects better what it describes.
2005-12-30 12:17:25 +00:00
ccremers
4a363aa33c - Bugfix: the single attack output adaptions from revision 1447 had
broken some of the xml output.
2005-12-30 12:03:19 +00:00
ccremers
ebf50b5252 - Removed the bold for complete proof. 2005-12-29 13:36:01 +00:00
ccremers
3b897c3872 - Added '--check' switch, to see whether your protocol terminates at all
if there is no intruder.
- Restructered many switches.
2005-12-29 12:52:51 +00:00
ccremers
a50245734d - Fixed the broken '--no-intruder' switch. 2005-12-29 12:14:21 +00:00
ccremers
c79c9eb73f - Added color output but forgot to add the sources files, fixed.
- Note: ~ is not expanded in SCYTHERDIR because it is not handled by the
  shell; thus $HOME should be used.
2005-12-29 11:03:18 +00:00
ccremers
515dec7f8b - Added note on usage of '~' in SCYTHERDIR. 2005-12-29 10:56:09 +00:00
ccremers
ab2f2469c0 - Added help text for the environment variables. 2005-12-29 09:35:08 +00:00
ccremers
1aca8460a3 - SCYTHERDIR environment variable is now used: colon-separated list of
directories to search. Oblivious to trailing slashes.
2005-12-29 09:25:42 +00:00
ccremers
efb3ec232b - Added some new thoughts, cleaned up old ones. 2005-12-28 22:52:22 +00:00
ccremers
d21f292330 - Renamed '--monochrome' to '--plain', which is nicer and shorter. 2005-12-28 18:40:58 +00:00
ccremers
3686a69869 - Added SCYTHERFLAGS environment variable. 2005-12-28 16:33:08 +00:00
ccremers
ab75acea62 - Added colour output, with --monochrome switch to disable this. 2005-12-28 15:27:22 +00:00
ccremers
e19f8bddd1 - Improved Reachable claims output
- Use square brackets for remark output instead of normal brackets.
2005-12-28 14:42:46 +00:00
ccremers
bceaca28f0 - Improved some type of warnings for e.g. empty prec sets andsoforth. 2005-12-28 14:25:06 +00:00
ccremers
72162e82c6 - Some cleanup. 2005-12-28 12:13:17 +00:00
ccremers
39adf85c6a - Reverted previous version: Tickets can possibly be secret as well
(although this is dubious, as the claiming role will not know the
  contents in many cases)
2005-12-28 12:04:00 +00:00
ccremers
dce2befd50 - Don't add secrecy claims for ticket types. 2005-12-28 11:59:39 +00:00
ccremers
0505aaacd6 - New claim: CLAIM_Reachable
- Added new switches:
  -G,--generate-statespace
  -C,--generate-claims
- Claims are now allowed to have no label (they will be generated
  automatically)
- Output summary shows parameter of claims
- Internally, new symbols can now be generated by
  symbolNextFree(prefixsymbol)
2005-12-28 11:50:17 +00:00
ccremers
ccc4c34823 - Added '--remove-claims' switch to cut off all existing claims. 2005-12-27 13:53:49 +00:00
ccremers
ca4c5674ac - Added check for non-used variables. 2005-12-27 13:44:12 +00:00
ccremers
397298290b - Improved output significantly. 2005-12-27 12:24:12 +00:00
ccremers
5ff71fa661 - Some cleanup.
- Added 'all-atacks' switch.
2005-12-27 12:01:17 +00:00
ccremers
28774cb94c - Moved dot output (finally) into a separate file, and made some minor
improvements.
2005-12-27 11:50:46 +00:00
ccremers
c4628e8be6 - Added support for more intelligent bounding. Fairly untested at the
moment.
2005-12-27 11:19:45 +00:00
ccremers
c20810def5 - Added preliminary support for singular attack output. 2005-12-27 10:49:22 +00:00
ccremers
ff503b24af - Added some todo stuff for the very near future. 2005-12-26 16:45:16 +00:00
ccremers
cb2aef3915 - Old state/time info has now been removed. This was only needed for the
POR engine anyway, so that's where it is shown now.
2005-12-26 16:28:45 +00:00
ccremers
32f226f782 - Better switches explanation. 2005-12-22 12:33:35 +00:00
ccremers
aae3cd70cb - Lowered default number of runs to 5. 2005-12-22 12:27:34 +00:00
ccremers
0259b2302c - Changed default behaviour: -d is needed to get dot output. 2005-12-22 12:24:27 +00:00
ccremers
e82ce8b962 - Added --no-intruder switch, but it is currently broken. 2005-12-21 19:02:41 +00:00
ccremers
aab5328a9b - Added a note. 2005-12-09 13:15:34 +00:00
ccremers
6543a8f659 - Added '--extravert' switch, which avoids initiator Alice to talk to
Alice.
2005-11-29 09:15:16 +00:00
ccremers
5276630007 - FIX: DOS newlines are now also accepted. 2005-11-28 08:27:05 +00:00
ccremers
e51b54af23 - FIX: Instantiation of variables is now the default.
- NEW: -C --class switch to reset this.
- NEW: max runs is now 6 by default for usability. For unbounded search,
  use -r 0 or --maxruns=0
2005-11-12 21:26:50 +00:00
ccremers
41132afea3 - Finally fixed the 'IV', 'RV' nuissance for global variables such as
the role names.
2005-11-12 21:16:02 +00:00
ccremers
76666404b0 - Added '--concrete' switch to fill in to pick readable names for
variables.
2005-11-12 21:13:00 +00:00
ccremers
c1c0b856de CHG: Changed default behaviour to Arachne engine.
NEW: Added 'S' switch for --summary things.
2005-11-04 13:23:30 +00:00
ccremers
5e1ca56f87 - Added experimental feature: explicit unique origination. This has to
be investigated further, because it seems to reduce just a few states.
  Note to Gijs: stay away from this, you should be writing your thesis.
2005-10-08 20:57:39 +00:00
ccremers
9f8f04c41c - Switched to the new consistency checking base. 2005-10-08 20:47:31 +00:00
ccremers
2ead7ab2ff - Improved code for new preferred order swapping of substitutions. Also
included more comments.
2005-10-08 20:22:24 +00:00
ccremers
2e495099bb - Added functions to globally check variable substitution consistency. 2005-10-08 20:03:31 +00:00
ccremers
2452a34671 - Added 'termlistMinusTermlist' function.
- Added TERMLISTERROR constant, and corresponding tests.
  Note that this will not work in many contexts, because only NULL is
  usually considered to be a special value. It is purely intended for
  the new type evaluation functions in type.c.
2005-10-08 19:56:04 +00:00
ccremers
fe730716ca - Added 'agenttypecheck' switch. 2005-10-08 19:54:30 +00:00
ccremers
a3b009f119 - The state of sys->variables was not maintained correctly, because
terms were destroyed before it could be tested whether they were in
  sys->variables. Thus, garbage was left in sys->variables. This has
  gone undetected because it was never really used. Hmpf.
2005-10-08 19:53:10 +00:00
ccremers
ea5bc6893f - Slightly smarter substitution in the symmetric case, when we make Var1
equivalent to Var2.
2005-10-07 20:38:41 +00:00
ccremers
83bf0ec704 - Added tuple_to_termlist function. 2005-10-07 14:02:46 +00:00
ccremers
1c075db161 - Added '--experimental=<int>' switch for debug version. 2005-10-07 13:09:07 +00:00
ccremers
5b73d707a0 - Rewrite of actor/agent type consitency code: now more aware of
initiator/responder difference.
2005-09-09 10:05:29 +00:00
ccremers
5c0c5d3333 - Better XML output for variables section. 2005-09-08 13:30:00 +00:00
ccremers
01a45f87d2 - Rewrote variables section, integrated typeflaw info.
Note that for variables that are not instantiated, only the variable
  and the type info is shown. For instantiated variables, both are
  shown.
  In this output the attribute 'free' should be ignored, as its
  output is not accurate here.
2005-09-08 12:55:32 +00:00
ccremers
48574de13e - Minor improvement to type flaw output: now also shows the type of the
substituted term.
2005-09-07 07:21:13 +00:00
ccremers
e104dddbfb - Added a switch to number the limit of intruder actions.
Initial testing suggests it does not influence the number of states
  much for values of 2 and higher.
2005-08-21 21:38:32 +00:00
ccremers
c330a5b719 - Added some thoughts. 2005-08-21 21:36:00 +00:00