Commit Graph

32376 Commits

Author SHA1 Message Date
Jay McCarthy
84fcd48039 stlc sub 2 2014-03-19 16:32:27 -06:00
Jay McCarthy
9b8697fa6e stlc-sub 1 2014-03-19 16:32:27 -06:00
Jay McCarthy
6e3a9861c6 removing fixed 2014-03-19 16:32:27 -06:00
Jay McCarthy
c1db657163 stlc 9 2014-03-19 16:32:27 -06:00
Jay McCarthy
0d8dab679f stlc 8 2014-03-19 16:32:27 -06:00
Jay McCarthy
253e8ee669 stlc 7 2014-03-19 16:32:27 -06:00
Jay McCarthy
d6dc594662 stlc 6 2014-03-19 16:32:27 -06:00
Jay McCarthy
0d14e14b21 stlc 5 2014-03-19 16:32:27 -06:00
Jay McCarthy
e31703fc35 stlc 4 2014-03-19 16:32:26 -06:00
Jay McCarthy
a4f3585101 stlc 3 2014-03-19 16:32:26 -06:00
Jay McCarthy
f9946f4b95 stlc 2 2014-03-19 16:32:26 -06:00
Jay McCarthy
7fc57ba1ad stlc 1 counter 2014-03-19 16:32:26 -06:00
Jay McCarthy
33ca8d05db Fix diffs 2014-03-19 16:32:26 -06:00
Jay McCarthy
f7d554d727 Fix diffs 2014-03-19 16:32:26 -06:00
Jay McCarthy
0bdf7260ed Repair check and remove fixed 2014-03-19 16:32:26 -06:00
Jay McCarthy
9c4dcf2ea4 Ignore rejected patches 2014-03-19 16:32:26 -06:00
Jay McCarthy
01846909f9 Fix this screw-up 2014-03-19 16:32:26 -06:00
Burke Fetscher
dfe144e3b8 add bug6 to verifier benchmark 2014-03-19 16:23:58 -05:00
Robby Findler
67d16cbae5 make the list/e combinator be more fair 2014-03-19 15:28:59 -05:00
Robby Findler
09d78f0604 add counter examples to rbtrees 2014-03-19 14:16:28 -05:00
Robby Findler
a451fcfa47 fix performance bug in Redex
The bug was in the way matching worked for lists. Specifically, if
you define a grammar like this one:

  e ::= (- e)
        (- e e)
        integer

and you have a term like this:

  (- (- (- (- (- (- (- 11)))))))

then at each step of matching against 'e', Redex would try both
alternatives, meaning it is attempting 2^n matches (where n is
the number of nested minus signs).

The fix eagerly checks the lengths of the lists and so brings this
back to a linear time matching problem.

(This came up in the delimited continuation model from the paper
 _Constraining Delimited Control with Contracts_, ESOP 2013, altho
 Redex's caching can mask the bad behavior, making this linear
 again in normal uses of that model.)
2014-03-19 13:02:32 -05:00
Robby Findler
98719cb5cf unbreak after rename of make-mutants.rkt 2014-03-19 13:02:05 -05:00
Max New
6ecb0b765b Add convenience enumerator approximation function 2014-03-19 13:02:04 -05:00
Burke Fetscher
190ef41661 add exit message with counts 2014-03-19 12:59:40 -05:00
Burke Fetscher
4612775a75 exception handling for vm model 2014-03-19 12:29:46 -05:00
Burke Fetscher
9bb9f2b150 don't refresh generator on timeouts to avoid looping timeouts 2014-03-19 12:24:48 -05:00
Robby Findler
4a3e80d22b rename make-mutants because these aren't mutants in the usual sense of the testing literature 2014-03-19 11:44:14 -05:00
Robby Findler
bc9729358a adjust check function for poly-stlc and put Jay's counter
example for bug #2 into the file
2014-03-19 11:41:06 -05:00
Robby Findler
3d175cba40 add + to poly-stlc 2014-03-19 11:14:30 -05:00
Robby Findler
cc2d547389 reapply the diffs to get the new p values set for enumeration-based random generation 2014-03-19 11:10:25 -05:00
Robby Findler
1ac1027c02 add tests to the poly-stlc model 2014-03-19 10:49:55 -05:00
Asumu Takikawa
6a80de3d5d Remove unused imports and a useless comment 2014-03-19 01:57:08 -04:00
Asumu Takikawa
35af4b7264 Simplify internal data for TR classes with kws
Use keywords instead of literals for communicating
internal information from the TR front-end macro to the
back-end.
2014-03-19 01:57:08 -04:00
Asumu Takikawa
a21a4a151a Handle init/init-field order correctly in TR
Also handles the case where init and init-field clauses
are interleaved, in which case they need to be interleaved
in the type.

Closes PR 14408
2014-03-19 01:57:08 -04:00
Burke Fetscher
d8d4d9712a redex: update virtual machine benchmark diffs 2014-03-18 23:29:33 -05:00
Burke Fetscher
89ad8cd062 adjust enum distributions 2014-03-18 23:17:15 -05:00
Burke Fetscher
b540cae4df redex: fix rvm ordered generation in n benchmark 2014-03-18 23:08:01 -05:00
Robby Findler
2d42507515 new slate of stlc-sub bugs 2014-03-18 22:39:58 -05:00
Robby Findler
deede439fe fix subst function
- its contract said it accepted M's as the third argument, but it
  worked only for v's
- remove two redundant cases
2014-03-18 22:39:58 -05:00
Robby Findler
5e275e41ef add missing value production
also, use redex-check? instead of redex-check in 2 places
2014-03-18 22:39:57 -05:00
Robby Findler
1ecd4c0563 change stlc-sub's 'check' function to not assume
its input is well-typed
2014-03-18 22:39:57 -05:00
Burke Fetscher
dea6e7c6e9 redex: add virtual machine model to benchmark 2014-03-18 17:13:38 -05:00
Robby Findler
9e41d393d5 unbreak stlc-sub test #9 2014-03-18 17:03:01 -05:00
Robby Findler
9ca793a1c8 add new stlc-sub bugs 2014-03-18 17:03:01 -05:00
Asumu Takikawa
da3bc2f266 Make TR class tests run successfully on their own
Closes PR 14409
2014-03-18 18:01:07 -04:00
Robby Findler
2dc333de74 Misc cleanups to stlc and stlc-sub benchmarks:
- add +
 - tweak definition of c to make more like Palka's dissertation
 - remove non-bugs from stlc-sub
 - share test suite between stlc and stlc-sub
2014-03-18 16:43:39 -05:00
Robby Findler
e106ec381f renumber poly-stlc bugs to start from 1 2014-03-18 16:34:17 -05:00
Robby Findler
d6cdc325d5 remove links synthesized information from drracket bug report 2014-03-18 15:31:49 -05:00
Sam Tobin-Hochstadt
b961b55bb2 Fix description of citet requirements. 2014-03-18 16:28:41 -04:00
Robby Findler
2598704dde add Michał Pałka's l 2014-03-18 14:52:45 -05:00