Vincent St-Amour
484fb029f1
Rephrase a missed optimization message.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
a319ba1058
Fix missed-optimization tests to also look for optimization messages.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
e8621f58ea
Fix optimization tests to also look for missed-optimization messages.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
28692786b3
Have optimizer test harness look only at log messages that came from the optimizer.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
641dd731ab
Fix optimizer test harness to use the new syntax of with-logging-to-port.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
09015722e3
Use with-logging-to-port from unstable.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
47f48c08ad
Revise with-logging-to-port to avoid the extra dummy log entry.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
a2afe2c285
Adapt the optimizer's test harness for the new logging strategy.
2011-06-17 18:18:01 -04:00
Vincent St-Amour
28cbad21b9
Add a directory for optimizer tests that we know fail.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
9ae4f82c0a
Sort optimizer logs by source location, to prepare for eventual user consumption.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
5a151420e4
Keep source location around for derived pair optimizations.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
b8b061fb39
Don't print duplicate log entries.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
147032e428
Extra test for precision loss.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
0432ea16cb
Add an extra test that checks for non-reporting of false positives.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
b598c4a1eb
close-call -> missed-optimization
2011-06-17 18:18:00 -04:00
Vincent St-Amour
c73bc6c918
Report unexpected complex types in arithmetic expressions.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
2a4a4e8184
Generalize reporting of wasted precision to arbitrary exact arithmetic expressions.
2011-06-17 18:18:00 -04:00
Vincent St-Amour
cd027109c9
Report when exact arithmetic is used inside float expressions, which may make the extra precision useless.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
fb36786c4d
Add failing test for nicer reporting of cascades of missed optimizations.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
5d4ef5c624
Add a test case for close-calls on all-real arithmetic expressions.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
89ca99210e
Report close calls for pair/mpair optimizations.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
526a84f10b
Add extra checks to avoid spurious close call reporting involving integer operations.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
95bf94577c
Show irritant for the currently detected close calls.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
0fab9e11ad
Revert "Fix optimizer expected log format."
...
This reverts commit bdf1073f17
.
Return that test to the new format, which it was originally written in.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
33962befd4
Made logging format more compact and less ambiguous.
2011-06-17 18:17:59 -04:00
Vincent St-Amour
7b2c6f5765
First attempt at logging unexpected Real types popping up.
2011-06-17 18:17:58 -04:00
Vincent St-Amour
2e0dd6467a
Add support for close-call tests to the TR test harness.
2011-06-17 18:17:58 -04:00
Robby Findler
f8f289adc0
fix random place tests
2011-06-18 05:15:36 +08:00
Sam Tobin-Hochstadt
55a25a825a
Disable this test.
2011-06-17 15:17:20 -04:00
Eric Dobson
0671945a8d
Moved tests for equality to unit tests
2011-06-17 15:17:20 -04:00
Eric Dobson
b326269133
Moved new TR symbol testing into unit tests.
2011-06-17 15:17:20 -04:00
Eric Dobson
3a13cf0e78
Moved TR string testing to unit tests.
2011-06-17 15:17:20 -04:00
Eric Dobson
54a12afb49
Added types and some tests for port operations
2011-06-17 15:17:19 -04:00
Eric Dobson
c77e906c7a
Added types and tests for equality operations.
2011-06-17 15:17:19 -04:00
Eric Dobson
2d152bac79
Added types and tests for raising and handling exceptions
2011-06-17 15:17:19 -04:00
Eric Dobson
4ba12dd20e
Added types and tests for symbol operations
2011-06-17 15:17:19 -04:00
Eric Dobson
ab23db7b5b
Added types and tests for string operations
2011-06-17 15:17:19 -04:00
Eli Barzilay
f6620d5754
Close the input port so that the file can be deleted on windows.
2011-06-17 13:19:24 -04:00
Matthew Flatt
7089a17ba2
fix `make-limited-input-port' limit tracking for committed peeks
...
and make the port thread-safe
2011-06-17 11:14:50 -06:00
Matthew Flatt
b1e47eba45
add `place-message-allowed?' and fix related problems
...
Closes PR 11983
2011-06-17 10:33:27 -06:00
Robby Findler
8e80af6f42
random testing for messages sent over place channels
...
related to PR 11983
2011-06-17 18:46:59 +08:00
Danny Yoo
d78fccb75b
Fixing the comment-at-end-of-file bug (PR#11982)
2011-06-16 21:37:48 -04:00
Matthew Flatt
523cec90c6
fix JITted multi-valued `begin0' in a single-value context
...
and generate simpler code for constrained cases, including
single-value contexts and ignored results
Closes PR 11979
2011-06-16 18:16:02 -06:00
Vincent St-Amour
26541ffbbd
Unify types and filter behavior of not and false?.
2011-06-16 18:08:44 -04:00
Stevie Strickland
477c5acb2a
Fix names in test to make them test-specific.
2011-06-16 16:51:38 -04:00
Matthew Flatt
56423f330e
`procedure-arity-includes?' reports #f for keyword-requiring procs
...
by default; a new optional argument restores the old behavior
(but the default behavior is consistent with the old docs and with
the vast majority of existing uses)
The implementation is ugly for performance reasons. A new primitive
`prop:arity-incomplete' property determines when to return #f for
`procedure-arity-includes?' in default mode. A nicer implementation
would be to redefine `procedure-arity-includes?' at the kw-proc level,
but the bytecode optimizer's and JIT's treatment of the built-in
`procedure-arity-includes?' is important. The implementation choice
could be revisited after cross-module inlining is implemented.
Closes PR 11978
2011-06-16 12:36:16 -06:00
Stevie Strickland
649fe2f276
Fix #%app bindings for `with-contract'-based contract forms.
...
Closes PR11975.
2011-06-16 14:25:01 -04:00
Matthew Flatt
a919062143
support progress evts in result of `make-limited-input-port'
...
and also adjust initial position in various port constructions
to use `file-position' instead of `port-next-location'
2011-06-16 09:18:47 -06:00
Matthew Flatt
83d002a9aa
fix interaction of `port-commit-peeked' and port positions
...
--- includes a small hack to the `make-input-port' protocol
to specially handle a byte-string result as "true"
from the `commit' proc
2011-06-16 08:17:14 -06:00
Sam Tobin-Hochstadt
eaa156554a
Add contracts for `Sequenceof' values.
2011-06-15 17:02:09 -04:00