Activity
From 08/26/2014 to 09/01/2014
09/01/2014
-
05:46 PM Bug #6592: test_call_double(DL::TestDL) fails on ARM HardFP
- Fails with gcc-4.8.2 also.
~~~
$ make check
CC = gcc
LD = ld
LDSHARED = gcc -shared
CFLAGS = -O3 -fno-fast-math -ggdb3 -Wall -Wextra -Wno-unused-parameter -Wno-parentheses -Wno-long-long -Wno-missing-field-initializers -W... -
01:24 PM Bug #6592: test_call_double(DL::TestDL) fails on ARM HardFP
- ping ... any update? Still an issue with ruby 2.2.0dev (2014-08-26 trunk 47288) [armv7hl-linux]
-
05:34 PM Bug #9976: ENV doesn't raise SecurityError except for aset and delete
- Backported into `ruby_2_1` branch at r47346.
-
- * hash.c (env_aset, env_has_key, env_assoc, env_has_value),
(env_rassoc, env_key): prohibit tainted strings if $SAFE is
non-zero. [Bug #9976]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@47346 b2dd... -
- 見すごしていました。 test_machine_stack_size のタイムアウトは r46499 で収まったとみていいんでしょうか?
-
- r45911, r45912, r45917, r45918 and r45919 were backported into `ruby_2_1` branch at r47345.
-
- * signal.c (rb_f_kill): directly enqueue an ignored signal to self,
except for SIGSEGV and SIGBUS. [ruby-dev:48203] [Bug #9820]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@47345 b2dd03c8-39d4-4d8f-98ff-823f... -
04:49 PM Feature #7517: Fixnum::MIN,MAX
- An obvious use case would be to make algorithms know how large an integer they can represent in Fixnum without overflowing to Bignum. This is strictly a Fixnum thing, too, so I think having the constants on Fixnum is exactly right. In JR...
-
12:30 PM Bug #10192 (Assigned): Install local versions of bundeld gems
-
- Trying to build recent version of Ruby in Fedora's builder, which has no access to the internet, the "make install" fails with error:
~~~
installing bundle gems: /usr/share/ruby/gems/ (build_info, cache, doc, extensions, gems, ... - 11:17 AM Revision e8473a67 (git): * 2014-09-01
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47344 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
11:16 AM Revision c5487618 (git): * ext/win32ole/win32ole.c (rbtime2vtdate): try to convert millisecond
- of Time object to millisecond of VT_DATE VARIANT.
* test/win32ole/test_win32ole_variant.rb
(test_conversion_time2date_with_msec): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47343 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Eric Wong wrote:
> shibata.hiroshi@gmail.com wrote:
> ...
It's VM(AWS)
> $ uname -a
> ...
```
[hsbt@chkbuild001 ~]$ cat /boot/config-3.10.53-56.140.amzn1.x86_64 | grep HZ
CONFIG_NO_HZ_COMMON=y
# CONFIG_HZ_PERIODIC is not set
CO... -
05:52 AM Bug #10191 (Closed): Possible memory leak using dup and setting an unassigned instance variable (Windows)
- Steps:
1. Run the following code (tested on Window 7):
~~~ruby
class Leaky
attr_accessor :v
def leak; d = dup; d.v = nil; end
end
l = Leaky.new
while true; l.leak; end
~~~
2. Watch the memory usage i... -
- Eric Wong wrote:
> Can we keep it out-of-tree and use it like rubyspec?
Good point.
I implemented ```make update-coverage``` task like update-rubyspec.
https://github.com/ruby/ruby/pull/708/files
> ...
simplecov can switch ter...
08/31/2014
-
11:56 PM Bug #10190 (Closed): Segmentation fault when Thread::Queue#push is called through dRuby
- The fix has been backported at patchlevel 170.
-
09:52 PM Bug #10190 (Closed): Segmentation fault when Thread::Queue#push is called through dRuby
- I found a segmentation fault when `Thread::Queue` is attached to a dRuby server as a front object and `#push` is called from a dRuby client.
Here are programs, a procedure and logs.
## server.rb
~~~ruby
require 'drb/drb'
require '... -
08:16 AM Bug #9961: TracePoint can skip c_return with rb_rescue()
- backported into `ruby_2_0_0` at r47342.
note:
* ruby 2.0.0にはa_call/a_returnがありません。
* ruby 2.0.0だとcoerceの冒頭でto_sが呼ばれるますね。2.1だけ違う？ - 08:14 AM Revision b10c3b6d (git): merge revision(s) 46465,46469,46484: [Backport #9961]
- * vm.c (rb_vm_rewind_cfp): add new function to rewind specified cfp
with invoking RUBY_EVENT_C_RETURN.
[Bug #9961]
* vm_core.h: ditto.
* eval.c (rb_protect): use it.
* eval.c (rb_rescue2): d... -
07:59 AM Revision 8bd2d31d (git): * lib/benchmark.rb: Fix a syntax error.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47341 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- backported into `ruby_2_0_0` at r47340.
(note that ruby 2.0.0 doesn't have a_call/a_return events.) - 07:55 AM Revision 3ed06c80 (git): merge revision(s) 46471: [Backport #9964]
- * compile.c (rb_iseq_compile_node): put start label of block after
trace (b_call).
[Bug #9964]
* test/ruby/test_settracefunc.rb: add a test.
added assert_consistent_call_return() method check call/r... -
- 2.0.0にはそもそも th->passed_bmethod_me がなくて、現在のtrunkのelse節に相当する部分がないんですけど、どうしたもんでしょうか？
-
- backported into `ruby_2_0_0` at r47339.
- 07:39 AM Revision f6838bc9 (git): merge revision(s) 46463: [Backport #9957]
- * vm_core.h: add VM_FRAME_MAGIC_RESCUE to recognize normal block or
rescue clause.
* vm.c (vm_exec): use VM_FRAME_MAGIC_RESCUE on at rescue/ensure.
* test/ruby/test_settracefunc.rb: should not invoke b_return ... -
- backported into `ruby_2_0_0` at r47338.
- 07:35 AM Revision 7b55994c (git): merge revision(s) 45953,45961: [Backport #9767]
- * lib/test/unit/parallel.rb: fix test-all parallel failure if a test
is skipped after raise.
DL::TestFunc#test_sinf is skipped after raise on mingw ruby.
But it causes Mashal.load failure due to undefined cl... -
- backported into `ruby_2_0_0` at r47337.
- 07:25 AM Revision 1f452734 (git): merge revision(s) 45827: [Backport #9618]
- * ext/pathname/lib/pathname.rb (cleanpath_aggressive): make all
separators File::SEPARATOR from File::ALT_SEPARATOR.
Reported by Daniel Rikowski.
Fixed by Nobuyoshi Nakada. [Bug #9618]
* ext/pathna... -
- backported into `ruby_2_0_0` at r47336.
- 07:23 AM Revision 7eb57e5a (git): merge revision(s) 45720: [Backport #9571]
- * lib/fileutils.rb (rmdir): rescue Errno::EEXIST in addition to
ENOTEMPTY (and ENOENT), because SUSv3 describes that "If the
directory is not an empty directory, rmdir() shall fail and set
errno to [EEXIST] ... -
- backported into `ruby_2_0_0` at r47335.
- 07:22 AM Revision 96fac49c (git): merge revision(s) 45144: [Backport #9544]
- * lib/resolv.rb (bind_random_port): Rescue EPERM for FreeBSD which
security.mac.portacl.port_high is changed.
See mac_portacl(4) for details.
Reported by Jakub Szafranski. [ruby-core:60917] [Bug #9544]
gi... -
- backported into `ruby_2_0_0` at r47334.
- 07:19 AM Revision 6ffd2d5d (git): merge revision(s) 46344: [Backport #9902]
- * re.c (match_aref): should not ignore name after NUL byte.
[ruby-dev:48275] [Bug #9902]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_0_0@47334 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- (reminder)
-
- backported into `ruby_2_0_0` at r47333.
- 07:11 AM Revision a102b685 (git): merge revision(s) 46368,46371: [Backport #9913]
- * ext/digest/digest.c (rb_digest_instance_equal):
fix #== for non-string arguments. [ruby-core:62967] [Bug #9913]
* test/digest/test_digest.rb: add test for above.
* ext/digest/digest.c (rb_digest_instance_equ... -
- backported into `ruby_2_0_0` at r47332.
- 06:56 AM Revision bea02d1b (git): merge revision(s) 46417,46418: [Backport #9939]
- * array.c (yield_indexed_values): extract from permute0(),
rpermute0(), and rcombinate0().
* array.c (rb_ary_combination): iterate on a shared copy, and use
array of indexes instead of array of chosen objects... -
- Backported into `ruby_2_0_0` at r47331.
- 06:30 AM Revision 1abe5c8d (git): merge revision(s) 46416: [Backport #9942]
- * array.c (rb_ary_permutation): `p` is the array of size `r`, as
commented at permute0(). since `n >= r` here, buffer overflow
never happened, just reduce unnecessary allocation though.
git-svn-id: svn+ssh://ci.rub... -
- * lib/benchmark.rb (BENCHMARK_CLOCK): Process::CLOCK_MONOTONIC_RAW
is not supported on old linux.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47330 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
05:26 AM Feature #8923: Frozen nil/true/false
- Matz, what do you think about it?
-
- * test/benchmark/test_benchmark.rb (test_realtime_output): use
assert_in_epsilon which compares in relative range, instead of
assert_in_delta which compares in absolute range.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47329 ... -
02:38 AM Feature #10189: Coverage report with Simplecov
- Can we keep it out-of-tree and use it like rubyspec?
I may also extract the terminal-friendly coverage output of yahns
into a standalone gem, its only a few lines:
http://bogomips.org/yahns.git/plain/test/covshow.rb
http://bog... -
- I ported [simplecov](https://github.com/colszowka/simplecov) into ruby core for measurement of test coverage.
see. https://github.com/ruby/ruby/pull/708
simplecov is MIT license. we can bundle it to ruby. Can I merge it?
08/30/2014
-
- nagachika00@gmail.com wrote:
> I don't have time to investigate this right now.
> And zlib has no maintainer according to
> https://bugs.ruby-lang.org/projects/ruby/wiki/MaintainersStdlib
> Are there anyone who can handle this?
H... -
11:53 PM Bug #10101 (Closed): Zlib::GzipReader produce different outputs for different methods applied - Applied in changeset r47327.
----------
zlib: GzipReader#rewind preserves ZSTREAM_FLAG_GZFILE
* ext/zlib/zlib.c (gzfile_reset): preserve ZSTREAM_FLAG_GZFILE
[Bug #10101]
* test/zlib/test_zlib.rb (test_rewind): test each_byte
We mus... -
10:03 PM Bug #10101: Zlib::GzipReader produce different outputs for different methods applied
`read` returns a string with external encoding. In your case it seems to be `UTF-8`. The encodings of the given `IO` object are ignored. Using `Zlib::GzipReader.open` doesn't work either, by the way. It still ignores the b`, but as a...- 11:53 PM Revision 3bfd894a (git): * 2014-08-31
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47328 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 11:53 PM Revision 3cee6a22 (git): zlib: GzipReader#rewind preserves ZSTREAM_FLAG_GZFILE
- * ext/zlib/zlib.c (gzfile_reset): preserve ZSTREAM_FLAG_GZFILE
[Bug #10101]
* test/zlib/test_zlib.rb (test_rewind): test each_byte
We must preserve the ZSTREAM_FLAG_GZFILE flag to prevent
zstream_detach_buffer from:
a) returning Qni... -
- Even if the use-cases aren't really convincing, I think the informational purpose alone is a good enough argument to add it. If you search for “ruby fixnum max” or something similar, you'll find the interest is there, but the solution(s)...
-
- Thank you ko1 for providing a patch to `ruby_2_1`. I can apply it cleanly.
Ant thank you eric for your confirmation.
Backported into `ruby_2_1` branch at r47326. -
- * gc.c: change full GC timing to keep lower memory usage.
Extend heap only at
(1) after major GC
or
(2) after several (two times, at current) minor GC
Details in https://bugs.ruby-lang.or... -
- Backported into `ruby_2_1` branch at r47325.
-
- * ext/win32ole/win32ole.c (ole_create_dcom): use the converted
result if the argument can be converted to a string, to get rid
of invalid access. Thanks to nobu. [ruby-dev:48467] [Bug #10127]
git-svn-id: svn+ssh://c... -
- Backported into `ruby_2_1` branch at r47324.
-
- * process.c (open): use UTF-8 version function to support
non-ascii path properly. [ruby-core:63185] [Bug #9946]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@47324 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- The patch looks good to me.
But rubygems has upstream repository.
https://github.com/rubygems/rubygems/
Please report to it first. -
- * internal.h (WARN_UNUSED_RESULT): warn unused result by gcc 3.4
or later.
* symbol.c: declare some functions with WARN_UNUSED_RESULT.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47323 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 10:30 AM Revision 765e4e2e (git): symbol.c (rb_sym2id): do not return garbage object
- The dynamic sym passed to rb_sym2id may be a garbage object
(as accounted for by dsymbol_check). This fixes an occasional
segfault in "make test-all" for me.
No need to backport, this is from the new symbol GC feature.
git-svn-id: svn... -
- shibata.hiroshi@gmail.com wrote:
> I investigate this on Amazon Linux 2014.3
Is this shared machine/VM? What kernel config/version?
Kernel config may be in /proc/config.* , otherwise distros may put
kernel config in /boot/confi... -
- After r47260, rubyci is sometimes failed. It seems to relate https://bugs.ruby-lang.org/issues/10165 .
* http://chkbuild001.hsbt.org/chkbuild/ruby-trunk/log/20140830T000025Z.fail.html.gz
* http://fb64b.rubyci.org/~chkbuild/ruby-tru... -
- 残念ながらautoconfの--without-*では引数を取れないのです。
-
- ごめんなさい。間違ってました。
--withext-out
じゃなくて
--without-ext
と書きたかったのでした。
--without-ext の方が --with-out-extより
単語の切れ目として自然だし英語として自然だからというのが根拠です。
-
01:55 AM Feature #10186: configure の with-out-ext オプション
- > ビルド対象から除外する拡張ライブラリを指定する configure のオプション
> ...
根拠が無いと比較できないのではないだろうか -
- * test/benchmark/test_benchmark.rb: extend sleep time, as 1ms is
too short on some environments.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47321 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * lib/benchmark.rb (BENCHMARK_CLOCK): prefer Process::CLOCK_MONOTONIC_RAW
if available to more accurate measure.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47320 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Missed patch 2/2
-
- Pretty trivial and low impact, but I figure we might as well reduce
memory footprint a few kilobytes where we can and let the reductions
accumulate. Feature #10185 has much bigger impact.
* [PATCH 1/2] iseq_inline_storage_entry: 24...
08/29/2014
-
- Patch #2:
Subject: [PATCH] iseq (rb_iseq_t): remove iseq->iseq field
It is unnecessary and reduces transient malloc/free during compile.
For "ruby -e exit", valgrind reports over 300K reduction in
overall allocations.
before... -
- running "ruby -rpp -e 'pp GC.stat'", a reduction in
malloc usage is shown:
before:
:malloc_increase=>118784,
:oldmalloc_increase=>1178736,
after:
:malloc_increase=>99832,
:oldmalloc_increase=>1031976,
(numbers from x86-... -
- ビルド対象から除外する拡張ライブラリを指定する configure のオプション
が --with-out-extになっています。
個人的には、--withext-out の方がいいと思います。
win32/configure.bat は --withext-out になっています。 -
09:41 PM Revision 66d3b975 (git): * ext/psych/lib/psych/visitors/yaml_tree.rb: fix NameError dumping and
- loading. Fixes GH #85. Thanks @brentdax for the patch!
* test/psych/test_exception.rb: test for fix
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47319 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- look like integers but have a newline. Fixes GH #189
* test/psych/test_string.rb: test for fix
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47318 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- should merge the hash in to the parent.
* test/psych/test_merge_keys.rb: test for change. Fixes GH #202
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47317 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 09:02 PM Revision 53220c49 (git): * 2014-08-30
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47316 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- should not be treated as merge keys.
* ext/psych/lib/psych/visitors/yaml_tree.rb: hashes with keys
containing "<<" should roundtrip.
* test/psych/test_merge_keys.rb: test for change. Fixes GH #203
git-svn-id: svn+ssh://ci.ruby-lang.or... -
06:36 PM Feature #10177: Hash#has_key? and Hash#has_value? should be deprecated
- I liked has_key? but if matz prefers the other way I am fine,
and I like that there is a consistent definition there.
Making stdlib / corelib more consistent would be great, and hopefully
gem-installable in a modular way too, so
... -
- The problem with object_class is that it is longer than self.class
Remember that object_id used to be called id in the past before that was renamed.
-
02:02 PM Feature #10183: An alternative name for method `class`
- Thomas Sawyer wrote:
> I once suggested `object_class` to go along with `object_id`.
> ...
What is `__class__`? -
01:40 PM Feature #10183: An alternative name for method `class`
- I once suggested `object_class` to go along with `object_id`.
Also, see https://bugs.ruby-lang.org/issues/6478 which talks about `BasicObject#__class__`. -
- The method `class` is special in that it always has to have an explicit receiver in order to avoid crash with the keyword `class`. But this is very inconvenient. I have seen so many
self.class
in codes. I propose that there sho... -
- Indeed. I have a similar error but different:
def !; end # => :!
/usr/lib/ruby/2.1.0/irb/init.rb:250:in `rc_file': private method `!' called for #<Proc:0x00000001a19338@/usr/lib/ruby/2.1.0/irb/init.rb:273> (NoMethodError)
... -
- When I input the following lines in irb, the irb terminates.
~~~
$ irb
irb(main):001:0> def !; end
=> :!
irb(main):002:0> !
/usr/local/lib/ruby/2.1.0/irb/input-method.rb:153:in `gets': private method `!' called for false:FalseCla... -
05:11 PM Bug #10167: Prime#include?(mod) hangs up
- It's difficult to delete `Prime.inlude?` method because this method is included by `include Enumerable`.
So, I overrided `Prime.include?` and `Prime.instance.include?`. Please review my patch. -
- (2014/08/29 10:20), nobu@ruby-lang.org wrote:
>> > I belive fstring codes in string.c.
> OK.
I like current code.
--
// SASADA Koichi at atdot dot net -
- Koichi Sasada wrote:
> I belive fstring codes in string.c.
OK.
~~~diff
diff --git i/eval.c w/eval.c
index f0a06fb..fbe17d0 100644
--- i/eval.c
+++ w/eval.c
@@ -52,7 +52,6 @@ ruby_setup(void)
Init_BareVM();
Init_h... -
- (2014/08/29 5:30), nobu@ruby-lang.org wrote:
>
> What about moving `fstring` stuffs to vm.c?
I belive fstring codes in string.c.
--
// SASADA Koichi at atdot dot net -
- nobu@ruby-lang.org wrote:
> What about moving `fstring` stuffs to vm.c?
Your patch looks good to me. ko1? -
- What about moving `fstring` stuffs to vm.c?
-
- (2014/08/29 4:24), Eric Wong wrote:
> OK, I made r47312. I kept the initialization changes since I think
> it was ugly to do lazy st_init_table (in case we need thread-safety
> in the future, using something like rculfhash (from Use... -
- ko1@atdot.net wrote:
> (2014/08/28 19:51), normalperson@yhbt.net wrote:
> > Cleanup in case MVM development proceeds.
>
> Now, mvm is stopping.
I am sad :*(
> I don't like to include vm_core.h only for such purpose. It ... -
- (2014/08/28 19:51), normalperson@yhbt.net wrote:
> Cleanup in case MVM development proceeds.
Now, mvm is stopping.
I don't like to include vm_core.h only for such purpose. It extends
dependency, taking long build time, and so o... -
- nobu@ruby-lang.org wrote:
> The dependency of string.o in common.mk will need `$(VM_CORE_H_INCLUDES)`.
Thanks, r47310.
I've been spoiled by automake :x - 06:30 AM Feature #10182 (Closed): [PATCH] string.c: move frozen_strings table to rb_vm_t
- Applied in changeset r47310.
----------
string.c: move frozen_strings table to rb_vm_t
Cleanup in case MVM development proceeds.
* string.c: remove static frozen_strings
* string.c (Init_frozen_strings): new function
* string.c (rb_fs... -
- The dependency of string.o in common.mk will need `$(VM_CORE_H_INCLUDES)`.
-
- nobu@ruby-lang.org wrote:
> I don't think it is possible to emulate `openat` family by FD in user space.
> So adding `rb_cloexec_open2()` is a bad idea, IMHO, not only its name.
Right, we cannot emulate `openat`; this needs kernel ... -
- Joel VanderWerf <joelvanderwerf@gmail.com> wrote:
> On 08/29/2014 01:21 AM, Eric Wong wrote:
> >Joel VanderWerf <joelvanderwerf@gmail.com> wrote:
> >>On 08/29/2014 12:55 AM, Eric Wong wrote:
> >>>Joel VanderWerf <joelvanderwerf@gmai... -
- I don't think it is possible to emulate `openat` family by FD in user space.
So adding `rb_cloexec_open2()` is a bad idea, IMHO, not only its name. -
- Joel VanderWerf <joelvanderwerf@gmail.com> wrote:
> On 08/29/2014 12:55 AM, Eric Wong wrote:
> >Joel VanderWerf <joelvanderwerf@gmail.com> wrote:
> >>On 08/28/2014 02:53 PM, Eric Wong wrote:
> >>>I like this feature.
> >>>
> >>>If... -
- Joel VanderWerf <joelvanderwerf@gmail.com> wrote:
> On 08/28/2014 02:53 PM, Eric Wong wrote:
> >I like this feature.
> >
> >If matz approves, I assume you also want to add other *at functions?
> >e.g. fstatat, renameat, unlinkat, m... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47314 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * test/rss/rss-assertions.rb: ditto.
* test/rss/test_maker_itunes.rb: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47313 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 07:22 AM Revision 343f8b02 (git): string.c: remove vm_core.h dependency [ruby-core:64627]
- * string.c: revert part of r47311, add rb_vm_fstring_table(),
remove vm_core.h dependency. [ruby-core:64627]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47312 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 06:30 AM Revision 2bb4c268 (git): * 2014-08-29
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47311 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 06:30 AM Revision a1e95636 (git): string.c: move frozen_strings table to rb_vm_t
- Cleanup in case MVM development proceeds.
* string.c: remove static frozen_strings
* string.c (Init_frozen_strings): new function
* string.c (rb_fstring): remove check for frozen strings,
use per-VM table
* string.c (rb_str_free): use... -
- Hello, Genadi. Thank you for your report.
And thank you shibata san for your investigation.
Maybe this was already fixed by backport for #9454, #9828.
I'll close this ticket. -
- Hello, yogesh. Thank you for your report.
Please show us command line options to configure.
And could you try to reproduce with trunk?
08/28/2014
-
- Cleanup in case MVM development proceeds.
-
- I like this feature.
If matz approves, I assume you also want to add other *at functions?
e.g. fstatat, renameat, unlinkat, mkdirat, etc. -
09:45 PM Feature #10181 (Open): New method File.openat()
- The purpose of the openat() function is to enable opening files in directories other than the current working directory without exposure to race conditions. Any part of the path of a file could be changed in parallel to a call to open(),...
-
08:14 PM Feature #10177: Hash#has_key? and Hash#has_value? should be deprecated
- @ Mr. Benoit Daloze
oh.. Yes, that's what I meant!
Thank you. -
12:29 PM Feature #10177: Hash#has_key? and Hash#has_value? should be deprecated
- I guess your second sentence is meant the other way round:
"And I agree with him, I supposed Hash#key? should be preferred over Hash#has_key?, " -
02:44 PM Misc #10178: refinements unactivated within refine block scope?
- I had posted some more code but remembered "send" doesn't apply yet! Sorry for my confusion. Any plans on indirect method access?
-
- Nobuyoshi Nakada wrote:
> In general, the scope inside a method definition is different from outside.
> ...
So I was correct, in that `refine` invokes a different scope where the refinements aren't activated? Ok, cool.
That's kind o... -
- In general, the scope inside a method definition is different from outside.
Consider method arguments and class/module level local variables. -
- Here is an executable version of what I was roughing out above, I apologize for not vetting it beforehand to prevent confusion:
~~~ruby
require 'active_support/core_ext'
module A
refine Time do
def weekday
self.strfti... -
- Nobuyoshi Nakada wrote:
> I can't get your point.
> ...
Yes, I mistakenly left out the "do" after `refine ActiveSupport::Time` (which should be `ActiveSupport::TimeWithZone`) and `refine Time`, with it the code does indeed work, and my... -
- I can't get your point.
`Module#refine` requires a block, so your code doesn't work, simply. -
- I doubt I am seeing a bug, but I was hoping someone could clarify for me the reason why I am seeing what I see. I tried pouring over the spec and wasn't quite able to pin it down.
My use case of refinements is not the normal one, so t... -
- * io.c (argf_next_argv): fix uninitialized variable in skipping
message when inplace edit without backup.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47309 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * test/ruby/test_io.rb: (test_invalid_advise): fix pattern for non
UTF-8 environment.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47308 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Actually, 1.9.3 is affected by this issue as well.
-
- This seems to be fixed by r46899, r46903 and r46904, backported for Ruby 2.1 by r46908, but it would be nice to have this also in Ruby 2.0, since that is the version officially shipped with RHEL/CentOS 7.
-
11:12 AM Bug #10180 (Rejected): #to_hash vs. #to_h
- Ruby often has two conversion methods for an object, e.g. #to_s and #to_str, #to_i and #to_int, #to_h and #to_hash.
The former is used for explicit conversion, the latter is used for implicit conversion (from an object with identical me... -
10:54 AM Bug #10180 (Rejected): #to_hash vs. #to_h
- The class Hash has a method try_convert, that is documented as "Try to convert obj into a hash, using #to_hash method." Surpisingly, the Array class doesn't have a method to_hash, but it does have a method to_h which converts it into a H...
-
- * test/ruby/test_io_m17n.rb (test_ignored_encoding_option): fix
leaked file descriptors by r47305.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47307 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * io.c (argf_next_argv): preserve encodings in warning messages.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47306 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * io.c (rb_io_extract_encoding_option): preserve encodings in
warning messages.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47305 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 05:55 AM Revision 7a4f7046 (git): * 2014-08-28
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47304 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * io.c (advice_arg_check): preserve encodings in warning messages.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47303 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * io.c (rb_io_s_new): preserve encodings in warning messages.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47302 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I'm not sure if this should be a bug or a new feature.
-
- shankga@gmail.com wrote:
> Alternatively Net::HTTP could just detect https urls and do this for you.
This sounds good. I cannot think of any downside... -
02:39 AM Feature #10179 (Open): Net::HTTP::Get.new("https://google.com").basic_auth(user_name, password) should throw exception stating the need to set use_ssl to true
- Alternatively Net::HTTP could just detect https urls and do this for you.
I suggest:
raise Exception("Accessing https without ssl\n url:#{URL}\n Please use http.use_ssl = true") -
-
- Related PR: [GH-706](https://github.com/ruby/ruby/pull/706)
08/27/2014
-
11:11 PM Feature #9867: Introduce each/else block into ruby
- Sorry for the delay.
I don't think I am able in specifying it in needed level of detail: (. If that's up to me, the issue can be closed. -
- I referred to this Matz's remark.
http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/43765
And I agree with him, I supposed `Hash#has_key?` should be preferred over `Hash#key?`,
so I replaced all of `Hash#has_key?` and... -
10:46 PM Feature #10176 (Open): Document how to perform net/http calls in parallel
- The typhoeus gem is a big source of Ruby VM crashes. It seems to me that the only reason people use this gem is because net/http doesn't document how to GET many URLs in parallel. I'd like to see a simple code example in the net/http R...
-
- We could’t use `Kernel#proc` if someone wanted to write code in the same way on both Ruby 1.8 and Ruby 1.9.
Now that maintenance of Ruby versions 1.8 end, there’s no reason to prefer `Proc.new` over `Kernel#proc` anymore.
https://w... -
-
-
- The patch can be applied cleanly to trunk (and 2.0.0/2.1).
It can be a new issue.
I move this ticket to ruby-trunk.
-
- Go ahead. We need to experiment in real use cases.
Matz. -
02:36 PM Revision c4ad649e (git): * lib/tempfile.rb: remove "require 'thread'". its features are no
- longer used.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47301 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
02:06 PM Revision 3644b94b (git): fix failed condition
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47300 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset r47299.
----------
* lib/drb/acl.rb: Removed meaningless #to_s methods in interpolation.
[Feature #10174][ruby-core:64584]
* lib/erb.rb: ditto.
* lib/observer.rb: ditto.
* lib/rake/invocation_chain.rb: ditto.
* li... -
- I found meaningless #to_s methods in interpolation like ... `"Failed to load #{filename}, #{e.to_s}"`
So I changed these to like ... `"Failed to load #{filename}, #{e}"` -
- [Feature #10174][ruby-core:64584]
* lib/erb.rb: ditto.
* lib/observer.rb: ditto.
* lib/rake/invocation_chain.rb: ditto.
* lib/rubygems/command_manager.rb: ditto.
* lib/rubygems/config_file.rb: ditto.
* lib/uri/common.rb: ditto.
git-svn-... -
- Applied in changeset r47298.
----------
* lib/drb/drb.rb: use attr_reader instead of Module#attr.
[Feature #10172][ruby-core:64582]
* lib/irb/ruby-token.rb: ditto.
* lib/net/telnet.rb: ditto.
* lib/rdoc/ruby_token.rb: ditto.
* lib/thw... -
- Module#attr is already deprecated.
But I found Module#attr without boolean argument (it means this identical to attr_reader) in some files in /lib, so I replaced Module#attr with Module#attr_reader -
- [Feature #10172][ruby-core:64582]
* lib/irb/ruby-token.rb: ditto.
* lib/net/telnet.rb: ditto.
* lib/rdoc/ruby_token.rb: ditto.
* lib/thwait.rb: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47298 b2dd03c8-39d4-4d8f-98ff-823fe... -
- of VT_DATE VARIANT to nsec of Time object.
* test/win32ole/test_win32ole_variant.rb
(test_conversion_dbl2date_with_msec): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47297 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I suppose 'empty ensure' like that ↓ has no meaning,
```
begin
r.document args
rescue Errno::EACCES => e
dirname = File.dirname e.message.split("-")[1].strip
raise Gem::FilePermissionError, dirname
rescue Interrupt => e
... -
- It is already fixed in upstream https://github.com/rubygems/rubygems/blob/master/lib/rubygems/doctor.rb#L108
-
- Ruby 2.1.0 deprecated the `File.exists?`
But I found `File#exists?` in `lib/rubygems/doctor.rb`, so I replaced `File#exists?` with `File#exist?`. -
- 2.1.2p207 seems working.
-
05:00 AM Bug #10170: Core dumped when using recursion in define_method
- Okay, ruby 2.1.2 gives everytime
`abort (core dumped)`
for
`A.new.a`
in pry, irb and `ruby -e 'class A; define_method :a do; a; end; end; A.new.a'`
-
- Could you try newer versions?
-
- Initial issue is wrong a little.
So if we have this
> ~~~ruby
> ...
when calling in irb
> ~~~ruby
> ...
or in pry
> ~~~ruby
> ...
gives
segmentation fault (core dumped)
But
`ruby -e 'class A; define_method :a do; a; end; e... -
- So if we have this
```ruby
class A
define_method :a do
a
end
end
```
and call `A.new.a` we got `SystemStackError`
but with alias
```ruby
class A
define_method :a do
a
end
end
```
when calling
... -
- @Hiroshi SHIBATA
I really appreciate you doing what I should do. -
- Bravo!
-
- mathn library is deprecated on trunk [Feature #10169]
-
- 1.9.3-p392 is too old. Please try latest version of ruby 1.9 or 2.0/2.1
-
- * parse.y (coverage): use rb_ary_tmp_new_fill() instead of setting
internal structure directly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47296 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * test/ruby/test_rational.rb: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47295 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
08/26/2014
-
- * compile.c (iseq_set_sequence): assigned in REALLOC_N, no further
assignment.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47294 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 11:00 PM Revision 3f9aa744 (git): * 2014-08-27
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47293 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 11:00 PM Revision 65f7aa19 (git): compile.c (iseq_set_sequence): check for multiplication overflow
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47292 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
08:22 PM Bug #10153: File.open block does not throw "No space left on device (Errno::ENOSPC)" if the data fits the buffer of IO.write
- What is the conclusion for this post?
-
- Seems intended to ignore `IOError` and `NoMethodError`.
-
- Applied in changeset r47288.
----------
io.c: do not swallow exceptions at end of block
* io.c (io_close): ignore only "closed stream" IOError and
NoMethodError, do not swallow other exceptions at the end of
block. [ruby-core:6446... -
- * test/ruby/test_rational.rb: cherry-picked working assertions from r47263.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47291 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r47290.
----------
lib/mathn.rb: remove built-in methods
* lib/mathn.rb (Fixnum#**, Bignum#**, Float#**, Rational#**):
remove as these are now built-in. [ruby-core:63973] [Bug #10086] -
- * lib/mathn.rb (Fixnum#**, Bignum#**, Float#**, Rational#**):
remove as these are now built-in. [ruby-core:63973] [Bug #10086]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47290 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Reported by Tsuyoshi Sawada.
https://bugs.ruby-lang.org/issues/10135#note-1
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47289 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * io.c (io_close): ignore only "closed stream" IOError and
NoMethodError, do not swallow other exceptions at the end of
block. [ruby-core:64463] [Bug #10153]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47288 b2dd03c8-39d4-4d8... -
- Tsuyoshi Sawada wrote:
>
> ...
This is another story.
Time#to_f is not accurate.
There are workaround, though:
~~~ruby
Time.at(1381089302.195).to_r.to_f
#=> 1381089302.195
~~~
-
- John Smart wrote:
> It seems that somewhere in the MRI, someone is using floating-point math to represent a Time.
You use floating point math in 1381089302195 / 1000.0 which is
1381089302.1949999332427978515625.
The Time instan... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47287 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47286 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47285 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Even `Prime.include?(4)` hangs up too.
This method seems useless. -
- workaround:
~~~ruby
enumerables = []
module_include_p = Module.method(:include?).unbind
ObjectSpace.each_object(Class) do |klass|
if module_include_p.bind(klass).call(Enumerable)
enumerables.push klass
end
end
~~~ -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47284 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47283 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
-
- Applied in changeset r47282.
----------
* lib/mathn.rb: mathn library is deprecated on ruby 2.2.
[Feature #10169][ruby-core:64553] -
- Approved. Organize transition path.
Matz.
-
- [Feature #10169][ruby-core:64553]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47282 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 12:25 AM Revision 2530d3e3 (git): * 2014-08-26
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47281 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47280 b2dd03c8-39d4-4d8f-98ff-823fe69b080e