suzanne.soy/racket
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
20,262 Commits 17 Branches 54 Tags 278 MiB
faef39cbc6
Commit Graph

2967 Commits

Author SHA1 Message Date
Kevin Tew
528948291d Fix creation of types and events after the master gc has been switched out 2011-02-03 13:14:45 -07:00
Matthew Flatt
309e1aec4f re-generate `configure' script 
Merge to 5.1, along with e91f243b64
 2011-02-02 16:10:40 -07:00
Matthew Flatt
9d204c01cf fix problem with allocation while holding a future lock 
Merge to 5.1
 2011-02-02 16:09:45 -07:00
Eli Barzilay
e91f243b64 Avoid compiling plot if the source directory is missing 2011-02-02 16:09:49 -05:00
Vincent St-Amour
62327c5f50 Fix make-->vector doc typo. 
Closes PR 11698.

Merge to 5.1.
 2011-02-02 14:00:14 -05:00
Matthew Flatt
9134aa4ee8 add missing mutex unlock in error path 
Merge to 5.1
 2011-02-01 19:37:02 -07:00
Vincent St-Amour
a0e6d67db2 Fix build when disabling single-precision floats. 2011-02-01 18:00:26 -05:00
Kevin Tew
9f51cb838f Fix used_pages count for place channel messages 2011-02-01 09:17:41 -07:00
Eli Barzilay
60bb91fd30 New Racket version 5.1.0.2. 2011-02-01 03:50:12 -05:00
Matthew Flatt
dfa36afa89 change `floating-point-bytes->real' to always produce a flonum 2011-01-31 19:52:50 -07:00
Matthew Flatt
ff0a591b7b change FFI `_float' to coerce to Racket double-precision flonum 2011-01-31 19:33:07 -07:00
Matthew Flatt
3ef32d915b make inexacts `eqv?' only when precision is the same 
plus some other small fixes
 2011-01-31 19:16:33 -07:00
Matthew Flatt
dc49e6a364 fix configure reporting of --disable-float 2011-01-31 19:16:07 -07:00
Matthew Flatt
fad0224259 fix mysterx for single-precision floats 2011-01-31 19:15:31 -07:00
Matthew Flatt
1e09a544a6 adjust _float' and _double' FFI types to accept any inexact real 
whether single-precision or double-precision
 2011-01-31 18:04:05 -07:00
Matthew Flatt
f9bcdc9baf move float config to "mzconfig.h"; enable by default for Windows 2011-01-31 17:51:09 -07:00
Vincent St-Amour
e65b206e6e Add functions to convert back and forth between floating-point representations. 2011-01-31 16:39:17 -05:00
Vincent St-Amour
03ec1ec501 Add a single-flonum? predicate to test for single-precision floats. 2011-01-31 16:39:17 -05:00
Vincent St-Amour
ac76d963b0 Enable single-precision floats by default. 2011-01-31 16:39:17 -05:00
Vincent St-Amour
c997ae139b Single-precision floats now print differently from their double-precision 
counterparts.
 2011-01-31 16:39:16 -05:00
Vincent St-Amour
d220e0a10b Add read syntax for single-precision float specials. 2011-01-31 16:39:16 -05:00
Matthew Flatt
dd5f0dfc80 fix `read-language' exn construction in an EOF case 
Closes PR 11683
 Merge to 5.1
 2011-01-31 06:49:13 -07:00
Matthew Flatt
bc5ab1e031 fix problems with `read-language' error reporting 2011-01-30 07:47:46 -06:00
Eli Barzilay
c945f5d27d New Racket version 5.1.0.1. 2011-01-30 03:50:12 -05:00
Matthew Flatt
0e169cfbb7 fix version number 2011-01-29 13:39:34 -06:00
Eli Barzilay
f190e3efec Post-release version for the v5.1 release 2011-01-29 13:04:18 -05:00
Matthew Flatt
b5bc288f67 Windows: fix `path-list-string->path-list' to throw away quotes 
--- which seems to be what the command shell does with quotes in
 PATH, at least; Microsoft isn't in the habit of documenting this
 sort of thing
 2011-01-25 20:30:33 -07:00
Eli Barzilay
1e2a6ffd60 New Racket version 5.0.99.7. 2011-01-25 03:50:12 -05:00
Matthew Flatt
5e3dbbc3f8 change gracket -z' to just use read-eval-print-loop' 
instead of `textual-read-eval-print-loop'
 2011-01-24 19:47:21 -07:00
Matthew Flatt
069a7c2b48 add current-get-interaction-input-port' and adjust racket/gui' 
so that GUI events are dispatched while a REPL is blocked on input
 2011-01-24 19:47:20 -07:00
Vincent St-Amour
dfac6641f7 Typo. 2011-01-24 17:00:55 -05:00
Matthew Flatt
f36db863bb refine `char-graphic?' to include all chars in L and N categories 
Closes PR 11664
 2011-01-24 07:08:11 -07:00
Matthew Flatt
0c193a599a fix uncaught-exception-handler handling of non-escaping escape handler 
and change the emergency error display handler to log and error
 Closes PR 11630
 2011-01-20 14:01:33 -07:00
Matthew Flatt
6b115ecf33 fix glib log-message construction 2011-01-18 11:02:46 -07:00
Matthew Flatt
07a7916142 fix string-ref' and bytes-ref' error message 
Closes PR 11064, since the problem that triggered the bad
 error message is also fixed
 2011-01-16 20:00:39 -07:00
Matthew Flatt
21eee45a48 fix `read-language' when input has only comments 2011-01-16 19:01:45 -07:00
Robby Findler
1f66d545c0 adjust build.bat to respect PLT_SETUP_OPTIONS 
closes PR 11629
 2011-01-15 19:22:44 -06:00
Matthew Flatt
6e65311702 fix scheduler interaction with atomic mode 2011-01-15 16:46:33 -07:00
Matthew Flatt
c2afc03b3b fix interaction of `continuation-mark-set-first' and prompts 
including a documentation fix
 2011-01-15 08:05:55 -07:00
Matthew Flatt
84237c2cf3 fix optimizer bug that can attempt infinite unrolling of loops 2011-01-13 12:48:44 -07:00
Matthew Flatt
79cd92859c disable user-specific setup on `make install' by default 
and add a `--enable-userspecific' flag for `configure'
 2011-01-13 10:59:20 -07:00
Matthew Flatt
be1046baa1 fix GC log printout 2011-01-10 06:43:27 -07:00
Matthew Flatt
ea51d32e9d fix subprocess checking when the process moves itself to a new group 
Closes PR 11200
 2011-01-08 10:29:32 -07:00
Vincent St-Amour
b32287fa2d Fixed a bug with complex numbers made from floats of differing precision. 
It used to be that constructing complex numbers from a float and a
double sometimes caused incorrect coercions.
 2011-01-05 14:42:31 -05:00
Vincent St-Amour
2bf83832b9 Fix zero? error message. Accepts any number, not just reals. 2011-01-05 14:42:31 -05:00
Matthew Flatt
dd0b41d974 linux: reset to extended-precision fp before calling pow() 2011-01-04 16:13:02 -07:00
Matthew Flatt
e88c2142e6 some intptr_t printf fixes 2011-01-04 16:12:12 -07:00
Matthew Flatt
13ddab969b enable ptherads by default for OpenBSD 2011-01-04 12:44:36 -07:00
Matthew Flatt
cd676fe81c another context-cache repair 2011-01-03 18:04:11 -07:00
Matthew Flatt
c5de056da5 another continuation-context caching repair 
because the previous repair essentially disabled caching
 2011-01-02 16:38:25 -07:00
Matthew Flatt
a226abec59 fix entry loss in continuation-context caching 2011-01-02 16:04:26 -07:00
Matthew Flatt
dbd8faee1b fix scheduler bug 
where the particular combination of events seems not
 to have been possible until the extra delim-cont
 for try-atomic hook was added
 2011-01-02 09:24:15 -07:00
Matthew Flatt
bdc0d2c862 like f57b431c2e (missed other case) 2011-01-01 08:22:17 -07:00
Eli Barzilay
7cac1f6ac2 New Racket version 5.0.99.6. 2011-01-01 03:50:21 -05:00
Matthew Flatt
44cf598298 win32 file locking 2010-12-31 14:40:39 -07:00
Matthew Flatt
55693e090f add port-try-file-lock?' and port-file-unlock' 2010-12-31 14:40:27 -07:00
Eli Barzilay
1f49e35b21 2010 -> 2011 2010-12-31 15:59:39 -05:00
Eli Barzilay
e2cc0942cc Attempt to make some ignore files more robust 2010-12-31 15:59:39 -05:00
Matthew Flatt
e094cf4e51 win64: fix external reference to thread-locals 2010-12-31 07:42:40 -07:00
Matthew Flatt
42794ee337 rollback win32 transaction experiment 2010-12-30 17:31:05 -07:00
Matthew Flatt
4b1ead5bfe win64: stubs for stack-unwind support 2010-12-30 10:48:15 -07:00
Matthew Flatt
678bfa7fca win32: fix transaction creation 2010-12-30 10:43:52 -07:00
Matthew Flatt
05046c07a8 win32: implement 'rename-file-or-directory' with transactions, if available 2010-12-30 09:30:28 -07:00
Matthew Flatt
f57b431c2e JIT: although `values' is inlined, don't count it as "simple" 2010-12-29 10:39:37 -07:00
Matthew Flatt
3b032893c1 don't JIT-inline `values' in a single-value position 2010-12-28 17:47:03 -07:00
Matthew Flatt
84a4f3f006 remove a forced GC (surely a bad idea) in the weak-hash implementation 2010-12-28 16:16:24 -07:00
Matthew Flatt
3e3ed33cbf detect and report a case of deadlock 2010-12-28 11:09:45 -07:00
Matthew Flatt
51ff346e2f clarify mutual exclusion of gtk logging cases 2010-12-27 17:43:36 -07:00
Matthew Flatt
705b11f2b8 a prompt-test fix like the previous one, but for post thunks 2010-12-27 15:24:33 -07:00
Matthew Flatt
0d3fbb11fa supress a prompt test that doesn't apply to composable continuations 2010-12-26 16:21:28 -06:00
Matthew Flatt
0fd9a76081 VS SGC build 2010-12-26 09:35:26 -07:00
Matthew Flatt
b404ef9970 fix JIT bug related to 'apply' 2010-12-26 09:35:25 -07:00
Matthew Flatt
1c479bb0fb SGC fixes for Win32 2010-12-26 09:35:23 -07:00
Matthew Flatt
76b475aa27 avoid compiler warnings 2010-12-26 06:47:53 -06:00
Matthew Flatt
4094bd33f0 fix allocation bug in logging-message construction 2010-12-26 06:47:52 -06:00
Matthew Flatt
f11e66fd71 fix sgcdebug configuration 2010-12-26 06:47:51 -06:00
Matthew Flatt
e77917db8c fix another continuation bug discovered by Casey's random testing 2010-12-25 17:12:33 -06:00
Matthew Flatt
e7b52bf10f fix ffi handling of non-GCable pointers with offsets 
Closes PR 11567
 2010-12-25 09:34:07 -06:00
Matthew Flatt
78b2e47541 fix more try-atomic problems 2010-12-21 18:45:36 -07:00
Matthew Flatt
391e7f486f fix signedness bug in GC 
which fixes memory accounting in particular
 2010-12-21 12:39:18 -06:00
Matthew Flatt
ad890077d0 JIT-inline `values' and make it synced (not "unsafe") for futures 2010-12-21 11:06:35 -06:00
Matthew Flatt
17c802c2e1 compiler: (let ([x (begin M ... N)]) P) => (begin M ... (let ([x N]) P)) 2010-12-21 11:06:35 -06:00
Eli Barzilay
3751452530 New Racket version 5.0.99.5. 2010-12-16 03:50:17 -05:00
Matthew Flatt
31906d6261 route glib logging to Racket logging 2010-12-15 12:25:45 -07:00
Matthew Flatt
7f5ac2b7f9 restore Win64 fix to FFI 
that I had improperly added to "foreign.c" instead of the source
 2010-12-14 07:53:19 -07:00
Matthew Flatt
3b4a1c4e94 fix some GC-related problems 
due to pointer fields in GC-allocated structures that do not refer
 to GC-allocated memory; if the corresponding memory is free()ed and
 then the GC later takes over the same page of memory, then bad
 things can happen; some of the problems were crashing the Windows
 port while running the 2htdp image tests
 2010-12-14 05:53:48 -07:00
Matthew Flatt
790a91e520 fix minor potential GC bugs and add some debugging support 2010-12-13 16:31:13 -07:00
Matthew Flatt
04d1397825 yet another PRIdPTR fix 2010-12-13 16:31:13 -07:00
Matthew Flatt
8bfb7de833 revert overzealous printf format conversion 2010-12-12 14:06:33 -07:00
Matthew Flatt
f3c695e3df fix intptr_t printf template for Windows 2010-12-12 08:50:26 -07:00
Matthew Flatt
526c700411 fix more intptr printfs 2010-12-12 08:35:15 -07:00
Matthew Flatt
f9fb041598 use inttypes.h and PTIdPTR 2010-12-12 08:24:20 -07:00
Matthew Flatt
2906a6d750 mzcom minor clean-up 2010-12-11 06:37:44 -07:00
Robby Findler
6c75bda403 fix shadowed var 2010-12-10 20:36:31 -06:00
Matthew Flatt
2ae9f14f9e fix exn raised in refab-struct read errors 
Closes PR 11530
 2010-12-10 18:33:44 -07:00
Matthew Flatt
2b4f1a6908 extend `sync/timeout' to allow a tail-position fail thunk for polling 2010-12-10 18:06:21 -07:00
Matthew Flatt
f4f2b7c533 fix bug in macro expander 2010-12-10 13:41:05 -07:00
Eli Barzilay
adb657fc5f Bump _WIN32_WINNT to 0x0501, since that's what the compiler recommends to do 
when it complains about a value that is too low.

This complaint happens when building with 2010 -- I resolved these
complaints (here and in mysterx) by adding the definition to the vcxproj
files, but the 3m compilation still needs it.

For the record, the error text is:

  C:\Program Files\Microsoft Visual Studio 10.0\VC\ATLMFC\INCLUDE\atlcore.h(35) : fatal error C1189: #error :  This file requires _WIN32_WINNT to be #defined at least to 0x0403. Value 0x0501 or higher is recommended.
 2010-12-10 03:54:02 -05:00
Eli Barzilay
a8bf338962 Attempt to fix "libffi.vcxproj". 
* Sorted out mess for output directories and commands.
* Minimized some redundant repetitions.
* Used consistent OutDir and IntDir directories.
 2010-12-10 03:19:01 -05:00
Eli Barzilay
aa58d1773d Make "gc2/make.rkt" look for "win32/release" directories. 
I made the 2010 project files use a consistent path of
  $(Platform)\$(Configuration)\
so make this script look for "win32/release" before just "release".  (It
would be better to do the same in the 2008 project files.)
 2010-12-10 02:20:36 -05:00