Commit Graph

39520 Commits

Author SHA1 Message Date
Matthew Flatt
abc5c9cdef cs: propagate CC, etc., to Chez Scheme configure 2018-10-21 06:38:12 -06:00
Matthew Flatt
51258bc77c cs: add support for -A 2018-10-21 05:59:00 -06:00
Matthew Flatt
de98529c13 cs configure: support SDK options
For example, enable a 32-bit Mac OS build.
2018-10-21 05:59:00 -06:00
Matthew Flatt
3591a685fc reorganize configure to better support racketcs
Rearrange the configure scripts so that it will be possible to build
RacketCS from a source distribution and have it installed in the right
place. Also, when building Racket3m just to bootstrap RacketCS, don't
install Racket3m.
2018-10-20 20:19:36 -06:00
Matthew Flatt
0f18df2c1a makefile: set up for cs client build 2018-10-20 20:19:36 -06:00
Matthew Flatt
18f7b96828 makefile: add way to build RacketCS as racket 2018-10-19 11:19:44 -06:00
Matthew Flatt
8f7a5b6dd5 Add Racket HISTORY.txt note for v7.1 2018-10-19 07:42:55 -06:00
Matthew Flatt
b1dfb61223 racocs exe: fill in implementation 2018-10-17 12:55:53 -06:00
Paulo Matos
703deedb66 Avoid integer overflow
A runtime error was detecting and overflow on e which this fixes.
Fixes part of #2314.
2018-10-17 07:57:39 -06:00
Matthew Flatt
eb6705be3c thread: fix futures in a non-main place 2018-10-17 07:41:38 -06:00
Matthew Flatt
65e464e626 cs: fix concurrency problems in prefabs & interned datums 2018-10-17 07:41:38 -06:00
Alexis King
86bb85931d units: Fix an improper use of syntax instead of quote-syntax 2018-10-16 18:23:13 -05:00
Alexis King
5fb75e9f82 Fix handling of the scopes of contracted signature bindings
fixes #1652
2018-10-16 16:52:45 -05:00
Matthew Flatt
8b22ebbfbe cify: fix more GC problems 2018-10-16 15:03:45 -06:00
Paulo Matos
6d7ab42b56 Fix fasl read fasl string.
Add a new test for the failing case.
2018-10-16 09:29:55 -06:00
Matthew Flatt
a7ae05a414 thread: fix place-channel ephemeron problem
Retains a strong link to a place-channel write end when there's at
least one waiting thread. This is symmetic to keeping a strong link to
the read end when the place-channel queue is non-empty. The change
repairs a problem building documentation with places in `racocs
setup`.
2018-10-16 09:29:55 -06:00
Paulo Matos
c78787a259 Simplify copy-port
Simplify copy-port by removing the several `(let loop...` constructs.
2018-10-16 09:26:41 -06:00
Paulo Matos
f85810ddb5 Fix reference to root-module-path 2018-10-16 09:25:03 -06:00
Matthew Flatt
fa6bd87bcd jit: fix nary unsafe-{fx,fl}{=,<,>,<=,>=}
Fix a incorrectly nested "tiny jump" mode that turned off the
surrounding "short jump" mode.

Closes #2313
2018-10-15 21:03:57 -06:00
Matthew Flatt
e9855fa64f cs: fix assignment to auto struct fields 2018-10-15 17:22:14 -06:00
Matthew Flatt
c347bc0c99 cs: clone ChezScheme repo with --depth 1 2018-10-15 17:22:11 -06:00
Matthew Flatt
f9a69105bc cs: small repairs for chaperones
The "struct.rktl" and "chaperone.rktl" tests now pass.
2018-10-15 08:52:13 -06:00
Matthew Flatt
7bb3750ebc cs: fix some structure and imperonsator problems 2018-10-15 08:52:13 -06:00
Matthew Flatt
2b39e61c4e cs, thread, io: fix problems with evt chaperones 2018-10-15 08:52:13 -06:00
Matthew Flatt
7ecb663a6f cs: fix procedure-name problems 2018-10-15 08:52:13 -06:00
Ryan Culpepper
2166c56179 fix typo in error message 2018-10-15 10:50:59 +02:00
Matthew Flatt
48f7082bdc cs: fix problems with procedures, arity, and impersonators 2018-10-13 13:06:06 -04:00
Matthew Flatt
2051bdfc6a schemify: don't incorrectly inline a recursive function 2018-10-13 13:06:06 -04:00
Matthew Flatt
9310b1e14c cs: recognize more Scheme errors
We'll want a better approach eventually.
2018-10-13 13:06:06 -04:00
Matthew Flatt
a7989924fa regexp: don't equate #rx and #px 2018-10-13 13:06:06 -04:00
Matthew Flatt
c43e42c903 cify: fix a GC problem 2018-10-13 13:06:06 -04:00
Robby Findler
b97811d4db improve error checking in the struct form of contract-out
closes #2303
2018-10-12 20:37:39 -04:00
Matthew Flatt
51e08c48f1 thread: repair for killing a thread blocked on itself 2018-10-12 05:56:39 -06:00
Matthew Flatt
e260aef958 cs: fix more printer problems exposed by "print.rktl" tests 2018-10-11 20:32:22 -06:00
Matthew Flatt
dbb74b5814 doc and test updates for a sandbox repair 2018-10-11 16:03:03 -06:00
Paulo Matos
ad7797e2cf Fix memory leak
Free env variable if it was allocated.
2018-10-11 09:34:42 -06:00
Paulo Matos
9020f5aac5 Clarify the use of foff as a base pointer in foreign malloc
Fixes #2285.
2018-10-11 09:34:03 -06:00
Matthew Flatt
b223ce471e docs: describe some limits of place scaling
Closes #2304
2018-10-11 09:32:21 -06:00
Matthew Flatt
959b8778d3 improve docs on version numbers
Closes #2309
2018-10-11 09:17:15 -06:00
Matthew Flatt
9f839c1161 expander: avoid retaining some syntax objects
During the expansion of an internal-definition contexts, don't
retain full forms, and instead keep just enough to perform
syntax tracking, if any.
2018-10-11 06:55:04 -06:00
Matthew Flatt
26f785e64e cs: fix some printer problems exposed by "print.rktl" tests 2018-10-11 06:01:07 -06:00
Matthew Flatt
d25058c94a cs: fix problems exposed by "read.rktl" tests 2018-10-11 06:01:07 -06:00
Matthew Flatt
694fe2e55c docs: correct description of custom-port special results
There no longer seems to be any path that uses the 0-arity option for
a read result from a custom input-port that is a function for a
"special" result. Probably, that path got lost when I added a `source`
optional argument to `read-char-or-special`. If so, that effect of the
change wasn't intentional, but let's just be happy that a
simplification is in place.
2018-10-11 06:01:07 -06:00
Matthew Flatt
16347b434d cs: fix raise-result-arity-error 2018-10-11 06:01:07 -06:00
Matthew Flatt
191d17461f hash-ref: adjust error message for bad failure thunk
Refines 2ef8d60cc6 to avoid characterizing the failure as a `(-> any)`
contract on `hash-ref`, since `hash-ref` doesn't enforce that contract
in general. Go back to an `exn:fail:contract:arity` error, but keep
the specialization of the error message to clarify that it's from
`hash-ref`. Also, bring RacketCS into sync.
2018-10-11 06:01:07 -06:00
Alexis King
15280640d4 Track disappeared define-syntaxes in internal definition contexts 2018-10-10 11:54:44 -05:00
Jordan Johnson
5376a4b409 Clarify match's app pattern behavior (#2306)
If I have understood the behavior of ```app``` correctly, the documentation needs to be clarified. Currently it implies that match expects a single result, and one could reasonably infer that the matcher will require that single result to satisfy all ```_pat```s given. The error message resulting from a reasonable-but-wrong pattern like ```(app string->number (? number?) x)``` is inscrutable, so a doc fix is important here.

I have also added an example that illustrates the use case that led me to submit this PR.
2018-10-10 09:27:21 -04:00
Matthew Flatt
2b9c0c4689 avoid using directory-exists? as part of a contract
Although a `directory-exists?` check is useful for providing better
error messages, it's fundentally a race condition, since an external
process can always remove a directory between the check and a use of
the directory. Because of that limitation of `directory-exists?`, we
normally avoid making it part of a contract. This commit adjust
937aa3cdb1 to follow that convention while preserving the helpful
check and documentation improvements.
2018-10-10 06:26:57 -06:00
Alexander McLin
6c0d193132 Fix references to raco setup and raco make in generate-stripped-directory doc. 2018-10-10 05:51:20 -06:00
Alexander McLin
af9cf01ea3 pkg: add missing option to documentation 2018-10-10 05:51:20 -06:00