Matthew Flatt
d1a942be63
net/git-commit: add a #:strict-links?
argument
...
Fail on checkout if it creates a symbolic link with an absolute
path or a relative path with "..".
Adjust `raco pkgs` to use `#:strict-links? #t`.
2015-10-02 09:38:27 -06:00
Matthew Flatt
23e2b806a2
git-checkout: document and test #:strict-links?
2015-10-02 09:34:46 -06:00
Matthew Flatt
37814da758
add tests for net/git-checkout
2015-10-02 09:07:28 -06:00
Matthew Flatt
42cf80815d
net/git-checkout: fix checkout of symlinks
2015-10-02 09:05:34 -06:00
Matthew Flatt
4caaf3e8b3
Use "https://" instead of "git://" for GitHub download
...
I originally chose "git://" because I thought "https://"
from GitHub didn't support `--depth 1`, but that does not seem
to be the case.
2015-10-02 08:35:28 -06:00
Matthew Flatt
ed07a5e176
fixup
2015-10-02 05:50:06 -06:00
Matthew Flatt
8c51d50cd2
raco pkg: improve Git repo-download caching
...
Drop the path and branch from a URL that references a Git
repository, so that a single download is used for multiple
packages in the same repo.
2015-10-01 22:04:24 -06:00
Matthew Flatt
acac7092c5
doc tweak
2015-10-01 21:30:46 -06:00
Matthew Flatt
7671c15b17
more GC callback shapes to support Mac OS X 10.11
2015-10-01 21:23:25 -06:00
Matthew Flatt
7abe38e763
adjust namespace-relative treatment of compiled require
...
Removing all original module context doesn't work, because it
doesn't distinguish between fragments of syntax that had the
"inside-edge" scope without the "outside-edge" scope.
Record the presence of the outside-edge scope by using the
root scope, and convert the root scope to the current namespace's
outside-edge scope on evaluation.
2015-09-30 11:32:16 -06:00
Matthew Flatt
6e80609998
macro expander: fix over-eager pruning of use-site scopes
...
The bug could cause
#lang racket/base
(define x 'outer)
(define-syntax-rule (def-and-use-m given-x)
(begin
(define-syntax-rule (m)
(let ()
(define given-x 'inner)
x))
(m)))
(def-and-use-m x)
to produce 'inner when it should produce 'outer.
Thanks to Brian Mastenbrook for pointing the problem and
providing examples.
2015-09-30 08:39:03 -06:00
Matthew Flatt
31549082e6
avoid crash when interrupting bytecode unmarshal
...
Interrupting bytecode unmarshal for syntax objects could leave
half-constructed values in a table that is intended to resolve graph
structure. Clear out work towards a graph construction when
interrupted.
The most common symptom of half-constructed syntax objects was a crash
after a Ctl-C during startup.
2015-09-28 14:38:12 -05:00
Sam Tobin-Hochstadt
dfef5b43fc
Add --write
option to fuzzer.
2015-09-23 15:49:18 -04:00
Sam Tobin-Hochstadt
57b4920234
Check that serialized procedures are the right kind of procedures.
...
Fixes http://drdr.racket-lang.org/31801/pkgs/racket-test/tests/racket/stress/fuzz.rkt
2015-09-23 15:49:18 -04:00
Sam Tobin-Hochstadt
d6ad89764b
Add a few more pair checks.
2015-09-23 15:49:18 -04:00
Sam Tobin-Hochstadt
083029fa6f
Add pair check in unmarshalling.
...
Closes #1070 .
2015-09-23 15:49:18 -04:00
Matthew Flatt
aaf098f203
fix potential GC problem while setting up a place channel
2015-09-22 16:45:47 -06:00
Robby Findler
5a33856802
change or/c so that it takes the first ho projection
...
whose first-order predicate accepts a value, instead
of requiring that there be exactly one
2015-09-22 09:05:34 -05:00
Robby Findler
45b635f707
don't set permissions on windows, for now
...
(at some point hopefully we'll figure out
what's the right behavior here)
This is related to pull request #1060
2015-09-22 09:00:55 -05:00
Spencer Florence
35a0f0c71c
make sure git checkout have correct perms
2015-09-21 14:17:51 -05:00
Matthew Flatt
003bca503f
avoid compiler warnings
2015-09-20 21:13:27 -06:00
Matthew Flatt
93efe503ad
fix misuse of comparison macro
2015-09-20 21:13:27 -06:00
Matthew Flatt
1d5b34f48b
auto-detect when strip -S
works for archives
...
Otherwise, don't try to strip archives such as "libracket3m.a".
2015-09-20 21:13:17 -06:00
Matthew Flatt
b37c07a280
Windows: generate "lib/msvc" content from cross-compile
2015-09-20 19:54:42 -06:00
Matthew Flatt
15797a7951
Windows: use same TLS strategy for both VC and MinGW in 64-bit mode
2015-09-20 19:17:15 -06:00
Matthew Flatt
8d43c73a0c
windows: use same scheme_setjmp() protocol with all compilers
2015-09-20 18:13:59 -06:00
Matthew Flatt
6dfc20d3ec
fix inferred-name propagation for internal-definition contexts
...
Set the name while checking for an immediate expansion when
no other forms follow.
2015-09-19 19:27:13 -06:00
Robby Findler
7ee29b0239
Revert permissions change because of appveyer error (that I
...
should have seen before pushing, sorry)
This reverts commit 47f36952d6
.
This reverts commit 47776f343d
.
2015-09-19 20:24:51 -05:00
Spencer Florence
47f36952d6
removed duplicate case
2015-09-19 20:21:50 -05:00
Spencer Florence
47776f343d
make sure git checkout have correct perms
2015-09-19 20:21:50 -05:00
Matthew Flatt
bcc65ac92e
GC: separate old-generation finalizers from now ones
...
Avoiding a traversals of old-generation finalizers can
save a couple of milliseconds for a minor GC in DrRacket.
2015-09-19 16:49:58 -06:00
Matthew Flatt
7ff1cf3619
avoid traversal of full page table when restoring mprotects
...
Overlooked this traversal at 50df879e79
.
2015-09-19 15:16:30 -06:00
Matthew Flatt
584920b3a6
GC: clarify page field overloadings
...
Rename fields in a page record and split some of them with `union` to
better document the intent of each field.
This change is intended to have no effect on the GC's behavior. One
tricky case is the line dropped around line 3542 of "newgc.c". That
line reset `scan_boundary` (formerly `previous_size`), which on the
surface is inconsistent with leving objects before the boundary
without `marked` bits set. However, that line is reachable only
when geneation-1 objects are being marked (objects newly moved
there would not be unmarked), in which case `san_boundary` should
already be reset.
2015-09-19 14:06:26 -06:00
Matthew Flatt
658bac7f52
note on catalogs and building from a tag
...
Closes #1058 .
2015-09-19 09:40:43 -06:00
Matthew Flatt
1ddaad8d58
xform: accomodate tokens like "10_1"
...
Those show up in the Mac OS X 10.11 SDK in `availability`
annotations.
Closes PR 15154
2015-09-19 09:19:43 -06:00
Matthew Flatt
9768f632a2
Disable static-library install for Windows or Mac OS X
...
Also, OS X `strip` needs `-S` for archives, in case a static-library
install is enabled.
2015-09-19 09:05:45 -06:00
Matthew Flatt
a75bdbf0a1
remove unused variable
2015-09-19 08:52:19 -06:00
Matthew Flatt
ee9d797906
fix file descriptor handling for poll() without epoll()
...
This problem could cause busy-waiting after a network connection
on Android, for example.
2015-09-19 08:16:30 -06:00
Matthew Flatt
fccd86d67d
configure: --enable-racket=auto
to simplify cross-compilation
...
Using `--enable-racket=auto` causes a Racket for the current platform
to be built in a "local" subdirectory of the build directory as
support for cross-compilation.
2015-09-18 21:56:20 -06:00
Matthew Flatt
a88d52bd34
configure: make --enable-libs
on by default
2015-09-18 21:06:00 -06:00
Matthew Flatt
1a48418844
infrastructure for raco ctool --c-mods <dest> --runtime <dir>
...
Make the runtime-file gatherer, which is normally used by `raco exec`,
work also for modules prepared for an executable that embeds Racket.
2015-09-18 18:55:22 -06:00
Matthew Flatt
a9015e5484
GC: remove unmaintained counters
2015-09-18 12:40:14 -06:00
Matthew Flatt
40f9467c07
count phantom bytes as regular allocation
...
The original idea was to count phantom bytes as "administrative
overhead", but issues discussed in #962 identified problems
with that idea. Finish shifting the accounting to treat
phantom bytes as payload allocation.
2015-09-18 12:26:32 -06:00
Matthew Flatt
4abedf63e8
raco setup: don't try to "sync" a doc that isn't pre-rendered
...
On my machine, the sync path was sometimes triggered for "htdp-ptr".
(I'm not sure why; progress toward determinsitic bytecode might
be relevant.) Adjust the trigger for sync mode to check that the
needed "provides.sxref" file exists.
2015-09-18 11:53:35 -06:00
Matthew Flatt
0b9cda5018
avoid undefined behavior in hashing inexacts
...
Casting a negative floating-point number to an unsigned integer
is not ok. Corece to a signed integer, first.
Thanks to John Regehr for help.
2015-09-18 09:49:04 -06:00
Juan Francisco Cantero Hurtado
b7bcd4f687
Remove extra ')'.
...
Found by cppcheck.
2015-09-17 14:12:15 -06:00
Juan Francisco Cantero Hurtado
38661481ed
Missing ')' in bool.c
...
Found by cppcheck.
2015-09-17 14:12:15 -06:00
Jack Firth
d305ceffe1
Refactor command-name
2015-09-17 14:11:50 -06:00
Jay McCarthy
02de19eb6f
who? me?
2015-09-17 15:02:16 -04:00
Jack Firth
c05db1ecf4
Refactor raco's tool listing code
2015-09-17 11:48:44 -06:00