Activity
From 06/02/2015 to 06/08/2015
06/08/2015
-
09:15 PM Bug #11236: inconsistent behavior using ** vs hash as method parameter
- This is expected behavior, keyword argument hashes only support symbol keys, not hash keys. I suppose the error message could be better, though.
-
09:05 PM Bug #11236 (Closed): inconsistent behavior using ** vs hash as method parameter
- Using ruby 2.2.2 I see this work:
~~~
[18] pry(main)> def test(cmd, opts={})
[18] pry(main)* puts cmd,opts
[18] pry(main)* end
=> :test
[19] pry(main)> test(:ads, :d => 6, :f => :gah, {a:3,b:4} => 3)
ads
{:d=>6, :f=>:gah, {... -
01:33 PM Bug #11233: Module#include, Module#prepend can accept no arguments - Jeremy Evans, you are right.
-
11:03 AM Bug #10015: Performance regression in Dir#[]
- Experiencing similar results with Adrien's benchmark:
With **ruby-2.2.2**: 2.9347980790189467
WIth **ruby-2.1.5**: 1.848318
OS X Yosemite 10.10.3, Journaled HFS+ file system, case insensitive.
-
06:11 AM Bug #11235 (Closed): [BUG] Segmentation fault
- コード
~~~ruby
n = []
(1..10**9).each do |i|
n << i
end
puts n.length
~~~
実行時のログ
~~~
>ruby a.rb
a.rb:3: [BUG] Segmentation fault
ruby 2.0.0p598 (2014-11-13) [x64-mingw32]
~~~
配列の要素数の上限を超えたため発生していると思いますが、
[BUG]という表示は... -
04:24 AM Bug #11232: rb_compile_string from parse.y segfaults
- Thanks for the tip!
I managed to fix the segfault with the following code:
https://github.com/kyrylo/fast_method_source/blob/d05a19993cccade40176c5f10b9008d79d2fc1e5/ext/fast_method_source/fast_method_source.c#L84-L85
I changed th... -
02:35 AM Bug #11226: Ripper fails on mswin64
- Koichi Sasada wrote:
> そして、中身を見ると、なんかやばい気がするエラーが...。手元でも動かして
> ...
ん、現状起きてる(ことがある)のって、4ヶ月前に報告した #10852 くらいだと思うんですが、これやばいのです？
06/07/2015
-
10:50 PM Feature #10672: Enable SSL on cache.ruby-lang.org
- Heya,
This is Doc from Fastly. Just wanted to let you know that if you send an email with a request to be added to our shared (subjectAltName) cert to "support at fastly dot com", and mention you're on the open source plan, you should... -
08:11 PM Revision 0d3591f2 (git): * gc.c (obj_info): print method id for T_IMEMO/ment.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50799 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- On 2015/06/07 23:21, usa@garbagecollect.jp wrote:
> 軽くショックだったのでいちおう言っておくと、http://rubyci.org/ にあります。
> vc10-x64ってやつがそうです。
すみません、一応そっちも、確認したつもりだったんですが、vc10-x64 が、見
たタイミングでは白背景（更新されていない）だったような気がして、今は更新
されてないのかと勘違いしていました。
でも、 http... -
- Koichi Sasada wrote:
> CI ってあるんでしたっけ。
軽くショックだったのでいちおう言っておくと、http://rubyci.org/ にあります。
vc10-x64ってやつがそうです。 -
06:53 PM Misc #11061 (Closed): doc/extension.ja.rdoc 査読依頼
-
- This isn't a bug, it a consequence of the fact that they take an arbitrary number of arguments:
~~~
Module.new.include(Module.new, Module.new, ...)
~~~
Changing it would break code like:
~~~
extensions = []
extensions << Mod... - 02:01 PM Bug #11233 (Closed): Module#include, Module#prepend can accept no arguments
- `# Module#include, Module#prepend can accept no arguments
Module.new.include
Module.new.prepend
` - 03:06 PM Revision b2dd608a (git): * 2015-06-08
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50798 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
03:06 PM Revision 3fe13558 (git): fix typo [ci skip]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50797 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
02:59 PM Bug #11234 (Closed): RubyVM::InstructionSequence.load で case が落ちる
- RubyVM::InstructionSequence.load を有効にしてテストしていますが、以下のコードで Segmentation fault になります。
~~~
str = <<EOS
def test_is()
cond = "x"
case cond
when "x"
p cond
end
end
EOS
ar = RubyVM::InstructionSequence.compile(str).to... -
01:59 PM Bug #11231: Should rb_alloc_tmp_buffer() be public API?
- Thanks for the explanations!
-
02:14 AM Bug #11231 (Closed): Should rb_alloc_tmp_buffer() be public API?
- Benoit Daloze wrote:
> `rb_alloc_tmp_buffer()` is declared in `include/ruby/intern.h`.
`include/ruby/ruby.h`.
> ...
Yes, exactly.
> Is there a way to hide such internal function?
> ...
It's similar to `alloca()`, the result sh... -
- Nobuyoshi Nakada wrote:
> Is `vparser` in `parser_compile_string()` GCed?
It looks like so. I'm not sure how to verify that exactly,
but when I disable the GC, it doesn't segfault anymore.
I have added a second example, with `GC.... -
- Is `vparser` in `parser_compile_string()` GCed?
-
- I'm using `rb_compile_string` in my project and invoke it in a loop.
Unfortunately, it segfaults for me.
I've set up a repository with a minimum reproducible case:
https://github.com/kyrylo/rb_compile_string_bug
The bug can be ob...
06/06/2015
- 10:11 PM Revision 3eb63250 (git): * Move test cases from test/ruby/test_complex.rb to test/test_cmath.rb
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50796 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
08:35 PM Bug #11162 (Closed): Update documentation for CMath library
- Applied in changeset r50793.
----------
* lib/cmath.rb: [DOC] Add docs [ci skip][Fix GH-909][Bug #11162]
Patch provided by @davydovanton - 08:35 PM Revision e0934947 (git): * lib/cmath.rb: [DOC] Documentation improvements [ci skip]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50795 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 08:35 PM Revision a8b34c37 (git): * 2015-06-07
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50794 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 08:35 PM Revision 843019f4 (git): * lib/cmath.rb: [DOC] Add docs [ci skip][Fix GH-909][Bug #11162]
- Patch provided by @davydovanton
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50793 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- `rb_alloc_tmp_buffer()` is declared in `include/ruby/intern.h`.
However it seems this is a fairly internal API as it deals with a `VALUE` pointer and has undocumented liveliness/GC behavior.
There was a try to specify it in
https:... -
- rb_struct_s_members() is declared in include/ruby/intern.h.
However it seems this is a fairly internal API as it returns an hidden Array.
For instance, there was a try to specify the behavior in
https://github.com/rubinius/rubiniu... -
- rb_method_entry_t.
r50728 changed sharing `def's to isolating `def's
on alias and so on. However, this change conflicts
future improvement plan. So I change back to sharing approach.
* method.h: move rb_method_definition_t::flags t... -
- 2nd arg.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50791 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- to make more clear.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50790 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- builddir/ext/ripper を消して試しても再現していたのですが、
中田さんの助言にしたがい、srcdir/ext/ripper を消してやり直してみたら、この症状が無くなりました。
というわけで、閉じます。 -
- Or maybe `STDIN.dup`?
-
-
02:51 AM Bug #11223: Rake SignalException: SIGSEGV
- It is a duplicate of Bug #10978 or #11089
-
- Couldn't you show small code to reproduce it?
-
03:06 AM Bug #11228: Classes Cannot be used in case Statements
- PS: The definition of `Class#===` allows you to write
case "A String instance"
when String
puts "flaw_demo: This is, where the control flow should be."
when Regexp
puts "flaw_demo: Hi Regexp!"
... -
- This is per spec, please check the definition of `Class#===`, since a `switch` is sugar for calls to `===`.
String === String # => false
-
- * class.c (ins_methods_push): suppress a signed and unsigned
comparison warning.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50789 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
06/05/2015
- 11:55 PM Revision 5d983e4f (git): test/socket/test_nonblock.rb: try to avoid EMSGSIZE
- We want to test the IO::WaitWritable behavior, so silently
discarding Errno::EMSGSIZE prevents the test from being
effective.
[ruby-core:69466]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50788 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
11:47 PM Feature #11229 (Closed): [PATCH] socket: allow exception-free nonblocking sendmsg/recvmsg
- As documented before, exceptions are expensive and IO::Wait*able are too
common in socket applications to be the exceptional case. Datagram
sockets deserve the same API which stream sockets are allowed with
read_nonblock and write_no... -
11:21 PM Bug #11228 (Rejected): Classes Cannot be used in case Statements
- ~~~
#!/usr/bin/env ruby
def flaw_demo
cl_selector="A String instance".class
case cl_selector
when String
puts "flaw_demo: This is, where the control flow should be."
when Regexp
puts "flaw_demo: Hi Regex... -
- to avoid indirect access overhead.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50787 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- shevegen@gmail.com wrote:
> `IO.new()` wants a number.
> ...
I don't see the point.
Why not use existing `STDIN`, `STDOUT`, `STDERR` (or `$stdin`, `$stdout`, `$stderr`)?
No need to allocate new `IO` objects. -
10:37 PM Feature #11227 (Feedback): May it be possible to add some symbols to IO.new() ?
- Hello.
`IO.new()` wants a number.
~~~ruby
io = IO.new(1)
~~~
Would it be possible or desirable to also allow symbols?
Such as:
~~~ruby
io = IO.new(:stdout)
~~~~
The `:stdout` seems to make it slightly easier for me ... - 10:21 PM Revision 5a67d8e2 (git): * ext/socket/ancdata.c (bsock_sendmsg_internal): drop redundant assignment
- We call memset to zero `mh' a few lines above anyways, making
explicitly zero-ing each field unnecessary.
Compilers optimize this code away anyways, but it still costs me
4 lines of my small terminal and even smaller attention span.
gi... -
- 今朝の trunk を mswin64 でビルドしたところ、
ripper でこんなエラーが出ていました。
```
Finished tests in 4.319549s, 56.7189 tests/s, 423.4238 assertions/s.
1) Failure:
TestRipper::ParserEvents#test_invalid_global_variable_name [C:/ko1/src/ruby/trunk/test/... -
04:12 PM Bug #11225 (Closed): r50776 + r50780 + r50781 changes behavior on systems without HAVE_STRUCT_MSGHDR_MSG_CONTROL
- Applied in changeset r50784.
----------
* ext/socket/ancdata.c (bsock_sendmsg_internal): all arguments are
parsed even on systems without HAVE_STRUCT_MSGHDR_MSG_CONTROL
to prevent SEGV caused by passing Qnil to RARRAY_LENINT and
t... -
- On systems without HAVE_STRUCT_MSGHDR_MSG_CONTROL such as Solaris 10,
when passing 3 or more arguments to rsock_bsock_sendmsg and rsock_bsock_sendmsg_nonblock that internally call bsock_sendmsg_internal,
r50775 or before: NotImplemen... -
- Applied in changeset r50784.
----------
* ext/socket/ancdata.c (bsock_sendmsg_internal): all arguments are
parsed even on systems without HAVE_STRUCT_MSGHDR_MSG_CONTROL
to prevent SEGV caused by passing Qnil to RARRAY_LENINT and
t... -
- On sparc Solaris10, during make test-all, the following SEGV occurs.
~~~
TestSocketNonblock#test_recvmsg_nonblock_error = 0.00 s = .
/XXXXXXXXXX-trunk-50782/test/socket/test_nonblock.rb:266: [BUG] Segmentation fault at 0x00000000000... - 04:11 PM Revision 6f586ac4 (git): * 2015-06-06
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50785 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- parsed even on systems without HAVE_STRUCT_MSGHDR_MSG_CONTROL
to prevent SEGV caused by passing Qnil to RARRAY_LENINT and
to preserve behavior before r50776.
[Bug #11224] [ruby-core:69468] [Bug #11225] [ruby-core:69469]
git-svn-i... -
- ビルドできるようになりました。ありがとうございます。
-
- r50782 にて bootstrap test は PASS all 1012 tests になりました。
かわりに、 make test-all 中に
test/socket/test_nonblock.rb:266: [BUG] Segmentation fault at 0x00000000000008
が出ていますが、これは同時期の別の変更のせいの気がします。
-
- ありがとうございます。
cast をしないようにしてみました。
ちょっと試してみて貰えないでしょうか。 -
- Applied in changeset r50782.
----------
* internal.h: move definition of rb_cref_t to method.h.
* eval_intern.h: move definition of rb_scope_visibility_t
to method.h.
* method.h: change rb_cref_t::scope_visi from VALUE to
rb_scope_v... -
- r50743 の変更を見ていると、
eval_intern.h 内で rb_cref_t を無理やり構造体 rb_scope_visibility_t にキャストしてから操作していますが、
これと、従来からある cref->flags へのビット操作関数 CREF_PUSHED_BY_EVAL_SET その他の操作が、
少なくともビッグエンディアン環境では、対象のビットが被るかもしれませんし、
それ以外の環境でも、キャストによる代入によって元のビットを上書きする... -
- def->flags.visi = 0 ということになります。
~~~
test_syntax.rb ..............................Fstderr output is not empty
bootstraptest.tmp.rb:4: [BUG] vm_defined: unreachable: 0
ruby 2.3.0dev (2015-06-05) [sparc64-solaris2.10]
... -
- from long to rb_method_visibility_t.
* class.c (ins_methods_i): catch up this fix.
* class.c (method_entry_i): cast to st_data_t instead of `long'.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50783 b2dd03c8-39d4-4d8f-98ff-823fe6... -
01:17 PM Bug #11223 (Closed): Rake SignalException: SIGSEGV
- I having this problem whenever I try to perform a rake on Cmock. Any guidance would be greatly appreciated.
~~~
Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.
C:\Users\qnguyen... -
- * eval_intern.h: move definition of rb_scope_visibility_t
to method.h.
* method.h: change rb_cref_t::scope_visi from VALUE to
rb_scope_visibility_t.
[Bug #11219]
* vm.c (vm_cref_new): accept rb_method_visibility_t directly.
* vm_in... -
09:36 AM Bug #11222 (Closed): make test-all 時の TESTS に -j オプションをつけるとファイルが多重ロードされる
- rubygems の upstream をマージしようと作業をしていたところ、TESTS="-j4" のようにパラレル実行のオプションを付けると rubygems のテストが落ちてしまう現象に遭遇しました。
https://github.com/ruby/ruby/pull/915
この問題自体は rubygems を修正することで対応したのですが、tenderlove にマージできるか相談した所
https://github.com/rubygems... - 08:28 AM Revision 7ac8fe74 (git): ext/socket/ancdata.c (bsock_sendmsg_internal): portability fix
- * ext/socket/ancdata.c (bsock_sendmsg_internal): avoid msg_control
ptr if msg_controllen is zero to fix portability problems.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50781 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
06:12 AM Bug #11221 (Closed): Fix indentation in Zlib documentation
- The documentation of the parameter window_bits for Zlib::Inflate.new misses two colons to respect the rdoc indentation.
See option (8..15) in http://ruby-doc.org/stdlib-2.2.0/libdoc/zlib/rdoc/Zlib/Inflate.html#method-c-new - 12:18 AM Revision 20dedbfd (git): ext/socket/ancdata.c (bsock_sendmsg_internal): fix build error from r50776
- Systems without HAVE_STRUCT_MSGHDR_MSG_CONTROL still need
controls to be nil for the check after argument parsing.
Note: not actually tested on a real system without msg_control
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50780 b...
06/04/2015
- 10:11 PM Revision cee9f4a4 (git): io.c: reduce size of :wait_*able code paths
- * io.c (sym_wait_readable, sym_wait_writable): declare
(io_getpartial): use sym_wait_readable
(io_write_nonblock): use sym_wait_writable
(Init_IO): initialize sym_wait_*able
On 32-bit x86:
text data bss dec hex ... - 09:45 PM Revision 5c461d1e (git): doc/extension.rdoc: note rb_get_kwargs changes keywords_hash [ci skip]
- [ruby-core:68507]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50778 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 08:53 PM Feature #11207 (Closed): [PATCH] ext/socket/ancdata.c (bsock_sendmsg_internal): use rb_scan_args
- Applied in changeset r50777.
----------
ChangeLog: add entry for r50776 [ci skip]
* ext/socket/ancdata.c (bsock_sendmsg_internal): use rb_scan_args
[ruby-core:69439] [Feature #11207] - 08:53 PM Revision 4dd6c370 (git): ChangeLog: add entry for r50776 [ci skip]
- * ext/socket/ancdata.c (bsock_sendmsg_internal): use rb_scan_args
[ruby-core:69439] [Feature #11207]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50777 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 08:50 PM Revision c4675473 (git): ext/socket/ancdata.c (bsock_sendmsg_internal): use rb_scan_args
- This should make the code more maintainable and make it easier to
add keyword argument support for "exception: false" in the future.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50776 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- of limit from 1.5 to 2.0. It is ad-hoc fix to solve test failure
in travis-ci.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50775 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- 報告ありがとう御座います。
r50773 で、もうちょっと詳しいメッセージを出すようにしてみました。
どうなるか、試して頂けないでしょうか。
-
- r50743以降、sparc Solaris 10 にて bootstrap test が失敗します。
（64ビット、32ビット共）
~~~
test_syntax.rb ..............................Fstderr output is not empty
bootstraptest.tmp.rb:4: [BUG] unreachable
ruby 2.3.0dev (2015-06-03) [... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50774 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50773 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- 報告ありがとうございます。
とりあえず、同じレイアウトの構造体を使うようにしてみましたが（ちょっとかっこ悪い）、いかがでしょうか。 -
- Applied in changeset r50772.
----------
* vm_method.c (rb_add_method_iseq): use intermediate struct to
avoid initializing struct with variables.
[Bug #11217]
* method.h: add a comment about it. -
- avoid initializing struct with variables.
[Bug #11217]
* method.h: add a comment about it.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50772 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 04:02 PM Revision d4e75213 (git): * 2015-06-05
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50771 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Also constify the following functions.
* rb_resolve_refined_method()
* rb_method_entry_with_refinements()
* rb_method_entry_without_refinements()
* rb_method_entry_copy()'s parameter.
* class.c: catch up this fix.
* vm_insnhelper... -
- strftimeにはマイクロ秒などでの出力を指定する、%6N, %9Nというフォーマットがあります。
一方で、パースを行うstrptimeにはそのような指定子が現在ありません。
そのようなものに対するニーズはぼちぼちあるようなので入れませんか。
http://answers.splunk.com/answers/1946/time-format-and-subseconds.html
https://twitter.com/nalsh/status/6064... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50769 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50768 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
01:22 PM Bug #11209: [PATCH] Fix for String#+ when subclassed
- You mean the original behaviour was intentional?
-
- I think it is a spec, and there must be more duplicate issues here but can't find them now.
-
- My apologies, just realised the previous patch file was in the wrong format. Uploaded a new patch.
-
- Usaku NAKAMURA wrote:
> ちなみに、rubyがrb_w32_(w)openをテキストモードで呼び出すのはかなりレアケースのはずです(ぼくにはすぐにはやり方が思いつかないレベル)。
あ、ごめんなさい、嘘でした。(バイナリモードでオープンしない限りテキストモードで呼んでた。あれー？)
お詫びの印にパッチつけときます。 -
- openにそういうオプション引数を追加するのが早いでしょうね。
ちなみに、rubyがrb_w32_(w)openをテキストモードで呼び出すのはかなりレアケースのはずです(ぼくにはすぐにはやり方が思いつかないレベル)。 -
- fluentdという、OSSのログコレクタがあるのですが、これには`in_tail`プラグインというものがあります。
これは、ログファイルを監視して、ログがファイルに追記されたらその分を読み込んでJSONにして他に流します。
このログファイルはfluentdではない他の誰か、ApacheだったりRailsだったりします。
この手のログファイルを書いていく流儀はいくつかありますが、ご存じの通りRailsやApacheは基本的に特定のパスにひたすら追記し続け、
... -
- Does it happen with recent versions?
-
01:52 AM Bug #11001: 2.2.1 Segmentation fault in reserve_stack() function.
- I also run into the bug too.
~~~
william@debianbox:~/talentlines/webui$ lsb_release -da
No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 7.7 (wheezy)
Release: 7.7
Codename: wheezy
william@debia... -
- Akira Tanaka wrote:
> Windows で ino を使えるようにするのがいけないというわけではないのですが、
> ...
なるほど、確かに使えそうです。
手元の現状のコードで使おうとすると結構いじらないといけないのがアレですが、大きく書き変える場合はtestのが良さそうですね。 -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50767 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50766 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- [fix GH-923]
* include/ruby/st.h: ditto.
* include/ruby/util.h: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50765 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50764 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- [fix GH-922]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50763 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 01:09 AM Feature #11190 (Closed): [PATCH] openssl: use io/nonblock instead of fcntl
- Applied in changeset r50762.
----------
ext/openssl/lib/openssl/ssl.rb: use io/nonblock instead of fcntl
[ruby-core:69382] [Feature #11190]
IO#nonblock= is easier-to-read, potentially more portable, and
avoids redundantly setting flags. - 01:09 AM Revision 14cbd61f (git): ext/openssl/lib/openssl/ssl.rb: use io/nonblock instead of fcntl
- [ruby-core:69382] [Feature #11190]
IO#nonblock= is easier-to-read, potentially more portable, and
avoids redundantly setting flags.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50762 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
06/03/2015
-
- * class.c (move_refined_method): catch up this fix.
* gc.c (mark_method_entry): ditto.
* vm_eval.c (vm_call0_body): ditto.
* vm_insnhelper.c (vm_call_method): ditto.
* vm_method.c: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/tr... -
- Applied in changeset r50760.
----------
dir.c: FNM_SHORTNAME
* dir.c (dirent_match): match short names only when FNM_SHORTNAME
flag is given, for the backward compatibility, and the new
behavior is often dangerous. [ruby-core:6943... -
- Seems good.
Thank you, nobu. Commit it, please. -
- https://github.com/ruby/ruby/compare/trunk...nobu:bug/11206-fnm_shortname
-
- * dir.c (dirent_match): match short names only when FNM_SHORTNAME
flag is given, for the backward compatibility, and the new
behavior is often dangerous. [ruby-core:69435] [Bug #11206]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/tr... - 08:57 PM Feature #11208 (Closed): [PATCH] variable.c: remove generic ivar support for special constants
- Applied in changeset r50759.
----------
ChangeLog: add missing reference
[ruby-core:69441] [Feature #11208] - 08:57 PM Revision 99487313 (git): ChangeLog: add missing reference
- [ruby-core:69441] [Feature #11208]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50759 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 08:53 PM Revision 152d36a7 (git): variable.c: remove generic ivar support for special constants
- Special constants are all frozen since [Feature #8923] and cannot
support ivars. Remove some unused code we had for supporting them.
* variable.c (special_generic_ivar): remove flag
(givar_i, rb_mark_generic_ivar_tbl): remove functio... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50757 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50756 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50755 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- passed argument.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50754 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 07:13 PM Revision 9ff020a0 (git): * 2015-06-04
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50753 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * class.c (clone_method): use renamed name.
* vm_insnhelper.c (rb_vm_rewrite_cref): do not use `node' in variable
names.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50752 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
05:11 PM Bug #9743 (Closed): memory leak in openssl ossl_pkey_verify leaks memory
- The failure has been fixed, so we can close this ticket.
-
- Solaris 10 上の古いfccにて、r50728以降、vm_method.c (vm.c から読み込まれる）のコンパイルに失敗します。
下記はr50750でのエラーメッセージです。
~~~
fcc -O2 -KV9 -KVIS2 -DRUBY_EXPORT -I/usr/local/64/lib/libffi-3.0.10/include -I/usr/local/64/include -I. -I.ext/include/sparc64-so... -
02:42 PM Feature #11215: pack/unpack for (u)intptr_t
- パッチの中に /* j for intptr_t */ というコメントがふたつありますが、ひとつは J ですかね。
-
- あ、言い忘れましたが、j/Jはperlでは
j A Perl internal signed integer value (IV).
J A Perl internal unsigned integer value (UV).
だそうです。( http://perldoc.perl.org/functions/pack.html )
これはRubyのVALUEとはちょっと意味合いが違いますし、必ずしもポインタ幅でもないので、
本... -
- 以前にも似たようなものを提案したような気がしないでもないですが、
WindowsのAPIを叩いてるとどうしても必要になるので改めて提案します。
pack/unpackのテンプレート文字として、新たに、
j: intptr_t, native endian
J: uintptr_t, native endian
を追加することを提案します。
他のプラットフォームではlongがこれと一致することが多いためにあまり問題に
ならないようです... -
- * eval_intern.h: move decl. of rb_scope_visibility_set() to method.h.
* load.c: catch up this fix.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50751 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Windows で ino を使えるようにするのがいけないというわけではないのですが、
ファイルが同一かどうか確かめるのには現在でも test("-", f1, f2) が使えないでしょうか。
```
% ruby -e '
File.write("z", "")
f = File.open("z")
p test("-", f, "z")
File.write("z.new", "")
File.rename("z.new", "z")
p te... -
- 現在WindowsではFile::Stat#inodeは常に0を返しますが、
例えばあるlogrotate的な運用の行われるファイルを監視・開きっぱなしにして、追記されれば差分を読み、
rotateされた場合は検出して最新のログを読む…というような場合に、rotateを検出する際にinodeは使われます。
WindowsにはGetFileInformationByHandle()というAPIで得られるBY_HANDLE_FILE_INFORMATION構造体の... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50750 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * ext/rbconfig/sizeof/extconf.rb: Check __int128.
* ext/rbconfig/sizeof/depend: sizes.c depends on
ext/rbconfig/sizeof/extconf.rb.
* template/sizes.c.tmpl: Detect check_sizeof.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@507... -
- Now, all `me` have `me->def`.
* proc.c (rb_method_entry_location): ditto.
* vm.c (rb_vm_check_redefinition_opt_method): ditto.
* vm.c (add_opt_method): ditto.
* vm_eval.c (vm_call0_body): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/... -
- * vm_core.h: use enum method_missing_reason for
rb_thread_t::method_missing_reason.
* vm_eval.c: catch up this fix.
* vm_insnhelper.c: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50747 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
07:21 AM Revision b1175728 (git): * vm.c: eagerly allocate `loading_table`. This eliminates the need to
- do NULL checks when looking up the `loading_table` hash.
https://github.com/ruby/ruby/pull/918
* load.c: remove various NULL checks
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50746 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * vm_insnhelper.c (vm_call_method): add LIKELY().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50745 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50744 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- `flag' contains several categories of attributes and it makes us
confusion (at least, I had confused).
* rb_method_visibility_t (flags::visi)
* NOEX_UNDEF -> METHOD_VISI_UNDEF = 0
* NOEX_PUBLIC -> METHOD_VISI_PUBLI...
06/02/2015
-
11:52 PM Bug #11214 (Closed): Cannot Get Correct Binding from inside of C Method
- I am trying to get the arguments passed into a method using the binding. This is possible using pure Ruby:
```ruby
class RubyBindingClass
def foo(arg = nil)
return binding
end
end
binding = RubyBindingClass.new.foo
... - 11:08 PM Revision 938eb46e (git): * ChangeLog: typo.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50742 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 11:07 PM Revision b06258f5 (git): * gem/bundled_gems: updated to test-unit 3.1.1 and minitest 5.7.0.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50741 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
09:42 PM Feature #10017: Add `Hash#fetch_values`
- Hi, please let me know if there's anything I can do to move this ticket forward.
-
- ko1@atdot.net wrote:
> So the following check should be redundant.
>
> > if (rb_obj_frozen_p(obj)) rb_error_frozen("object");
>
> It should be just
>
> > if (rb_special_const_p(obj)) rb_bug("...");
I will rely on rb_ch... -
- Good patch.
Before generic_ivar_set(), frozen check is already done.
So the following check should be redundant.
> if (rb_obj_frozen_p(obj)) rb_error_frozen("object");
It should be just
> ...
-
- variable.c: remove generic ivar support for special constants
Special constants are all frozen since [Feature #8923] and cannot
support ivars. Remove some unused code we had for supporting them.
Can somebody please sanity check t... -
- However it is difficult because we can't call `self.respond_to_missing?` directly because we should ignore `C1#respond_to_missing?`.
Options:
(1) Ignore this issue as a spec (or known issue)
(2) Do `super` (it has compatibility issu... -
- ```ruby
class C0
def method_missing *args
p [:method_missing] + args
end
def respond_to_missing? *args
true
end
end
class C1 < C0
def foo
super
defined?(super)
end
end
p C1.new.foo
```
T... -
08:03 PM Revision 4fc87d4c (git): * vm_insnhelper.c: fix a typo. [ci skip]
- check_resopnd_to_missing -> check_respond_to_missing
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50740 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50739 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- at defined?(func()).
* test/ruby/test_defined.rb: add a test for this fix.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50738 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- It's clear to me that there's no rationale behind the current behavior.
The question is broader than that.
First, it's not only for `String`:
~~~ruby
class MyArray < Array; end
x = MyArray.new([1,2,3])
x.first(2).class == x[0... -
- Applied in changeset r50737.
----------
* vm_insnhelper.c (vm_defined): skip respond_to_missing? when
a method is available.
[Bug #11211]
* test/ruby/test_defined.rb: add a test for this fix. -
- ```ruby
class C
def respond_to_missing? *args
p args
end
def existing_method; end
end
p defined?(C.new.existing_method)
```
This script prints
```
[:existing_method, false]
"method"
```
even though exsiti... -
- a method is available.
[Bug #11211]
* test/ruby/test_defined.rb: add a test for this fix.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50737 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- ```ruby
class C
def respond_to_missing? *args
true
end
def test
p defined?(non_exist_method())
end
end
C.new.test
```
This script prints only `nil`. It ignores `respond_to_missing?`.
It should be a bug.
-
- move instruction body to the vm_defined() function.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50736 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Path by @enebo.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50735 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
05:01 PM Feature #11210: IPAddr has no public method to get the current subnet mask
- It would be useful to expose @mask_addr. I have done this at times by reopening IPAddr and adding an attr_reader. Then the string form of the mask can be obtained with
```ruby
IPAddr.new(addr.mask_addr, addr.family).to_s
```
#mas... -
03:56 PM Feature #11210 (Closed): IPAddr has no public method to get the current subnet mask
- Both to_s and to_string omit the subnet mask of an IP address. The only way to query it via public methods is to manually parse the output of inspect.
-
04:28 PM Revision ea0bc53e (git): * vm_method.c (rb_method_definition_set): remove a double assignment.
- Coverity Scan found this bug.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50734 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 03:08 PM Revision 6697ade6 (git): * 2015-06-03
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50733 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50732 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- See #10845.
I also have a Github PR at https://github.com/ruby/ruby/pull/920.
-
- While making a r50728, iseqval is needed (to mark correctly),
but now just iseqptr is enough.
* class.c: catch up this fix.
* gc.c: ditto.
* proc.c: ditto.
* vm_method.c: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@5073... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50730 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- IMO, ruby should offer an option to users to select matching short names or not. (maybe File::FNM_SHORTNAME, or somehow.)
And, the default should not match short names. -
- For better or worse, extension names are something special things on Windows, so what about matching extension of long names, but not short names?
~~~diff
diff --git i/dir.c w/dir.c
index a6934bd..2b03a72 100644
--- i/dir.c
+++ w/... -
- Running these commands as the administrator would fix the problem:
1. `fsutil 8dot3name set 1`
2. `fsutil 8dot3name strip /s /f c:\` -
- Conflict with #10819.
Can these reach a compromise? -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50729 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset r50728.
----------
* method.h: make rb_method_entry_t a VALUE.
Motivation and new data structure are described in [Bug #11203].
This patch also solve the following issues.
* [Bug #11200] Memory leak of method ... -
- Applied in changeset r50728.
----------
* method.h: make rb_method_entry_t a VALUE.
Motivation and new data structure are described in [Bug #11203].
This patch also solve the following issues.
* [Bug #11200] Memory leak of method ... -
- Applied in changeset r50728.
----------
* method.h: make rb_method_entry_t a VALUE.
Motivation and new data structure are described in [Bug #11203].
This patch also solve the following issues.
* [Bug #11200] Memory leak of method ... -
- Motivation and new data structure are described in [Bug #11203].
This patch also solve the following issues.
* [Bug #11200] Memory leak of method entries
* [Bug #11046] __callee__ returns incorrect method name in orphan
... -
- This should make the code more maintainable and make it easy to add
keyword argument (to support "exception: false") in the future.
This unfortunately may cause extra allocations, but I don't
think it's worth worrying about since th... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50727 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * test/socket/test_nonblock.rb (test_sendmsg_nonblock_seqpacket):
OSX raises EPROTONOSUPPORT.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50726 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 01:47 AM Revision 18e77438 (git): test/socket/test_nonblock.rb: new test for sendmsg_nonblock
- sendmsg_nonblock was not tested on any of my systems due to the
common 64K limit. I also don't believe UDP sockets are at all
a useful candidate for sendmsg_nonblock testing since they
should never block on sending.
git-svn-id: svn+ssh... -
- Lee Jarvis wrote:
> Yes exactly. This will remove much of the boilerplate I mentioned above. However, the functionality should be exposed (I'm not sure if your code was suggesting that this is a documentable feature or just a workaround... - 12:08 AM Revision cb1affb4 (git): lib/benchmark.rb: just use Process::CLOCK_MONOTONIC
- Assume Process::CLOCK_MONOTONIC exists (possibly emulated) and
there is no need to bloat our code by defining a compatibility
constant.
* lib/benchmark.rb: just use Process::CLOCK_MONOTONIC
[ruby-core:69390]
git-svn-id: svn+ssh://ci....