Activity
From 09/06/2018 to 09/12/2018
09/12/2018
-
10:32 PM Bug #15050: GC after forking with fibers crashes
- @normalperson
Eric,
Sorry been busy with 'this is a bigger mess than I thought' kinds of things.
I just ran ruby-loco on r64706 'fiber: fix crash on GC after forking', and the build passed.
Thanks, Greg -
08:49 PM Bug #15050 (Closed): GC after forking with fibers crashes
- Applied in changeset trunk|r64703.
----------
share VM stack between threads and fibers if identical in size
ec->vm_stack is always allocated with malloc, so stack cache for
root fiber (thread stack) and non-root fibers can be shared a... -
- sameer.deshmukh93@gmail.com wrote:
> @hsbt what do you think about writing the docs within C API
> files like `ruby.h`, `intern.h` and `extension.h` itself so
> that ruby-doc.org can run doxygen on them to generate the
> relevant ... - 08:49 PM Revision 12409ad2 (git): fiber: fix crash on GC after forking
- Remove the remainder of ROOT_FIBER_CONTEXT use and unnecessary
differences between the root and non-root fiber. This makes
it easier to follow new root fiber at fork time.
Multiple sources of truth often leads to bugs, as in this case.... - 08:49 PM Revision d40694de (git): cont.c (fiber_memsize): do not rely on ROOT_FIBER_CONTEXT
- We can check if the fiber we're interested in is the
th->root_fiber for the owner thread, so there is no need to use
ROOT_FIBER_CONTEXT.
Note: there is no guarantee th->ec points to
&th->root_fiber->cont.saved_ec, thus vm::thread_memsiz... -
- Applied in changeset trunk|r64703.
----------
share VM stack between threads and fibers if identical in size
ec->vm_stack is always allocated with malloc, so stack cache for
root fiber (thread stack) and non-root fibers can be shared a... - 08:49 PM Revision 7ae24709 (git): cont.c (ec_set_vm_stack): avoid needless casting
- Am I missing something, here? Casting was totally unnecessary
and ugly...
[ruby-core:88929]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64704 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 08:49 PM Revision c99b9eb0 (git): share VM stack between threads and fibers if identical in size
- ec->vm_stack is always allocated with malloc, so stack cache for
root fiber (thread stack) and non-root fibers can be shared as
long as the size is the same. The purpose of this change is to
reduce dependencies on ROOT_FIBER_CONTEXT.
[... -
- * [Feature #13618] auto-fiber name, is "Thread::Coro" OK?
How should Mutex work between coroutines?
Queue/SizedQueue works, now (not sure about signal handler). -
- > It goes on top of my coro-r64610 branch posted at [ruby-core:88800]:
> https://public-inbox.org/ruby-core/20180901131301.5peghdyrtmks5mka@dcvr
>
> Several tests are skipped because Thread::Coro#stop?/#status/#[]=
> aren't impleme... -
06:43 PM Misc #15109 (Closed): Improve safe navigation operator's docs
- Reason: [current docs](http://ruby-doc.org/core-2.5.1/doc/syntax/calling_methods_rdoc.html) look this way (in "Receiver" section, one paragraph before last):
> You may use `&.` to designate a receiver, then `my_method` is not invoked ... - 04:05 PM Revision 2bd4c514 (git): * 2018-09-13
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64702 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
04:05 PM Revision c0f62450 (git): [DOC] aseq.each always return self [ci skip]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64701 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
02:37 PM Revision 351b8a32 (git): time.c split time_utc_or_local
- * time.c (time_utc_or_local): split into time_s_mkutc and
time_s_mktime without utc flag.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64700 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
01:32 PM Revision f7b0a791 (git): _mjit_compile_pc_and_sp.erb: forget about memcpy [ci skip]
- because using memcpy here actually made performance worse.
In Optcarrot,
* for loop
fps: 55.90428960597268
fps: 57.21690379342434
fps: 58.41751168279789
fps: 58.7317470613203
fps: 62.81371000202459
fps: 63.02541045417479
fps: 63.7732402... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64698 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 08:51 AM Revision 98fa0ab4 (git): [DOC] Modify descriptions for ArithmeticSequence
- [ci-skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64697 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 08:36 AM Revision 1595421c (git): [DOC] Add the documentation of ArithmeticSequence
- [ci-skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64696 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
07:48 AM Feature #14982: Improve namespace system in ruby to avoiding top-level names chaos
- Please allow me to humbly present [this design and implementation of import semantics](https://gitlab.com/lloeki/pak) aiming to solve precisely the current issue, taking inspiration from Go and Python.
Basically it implements a `Packa... - 07:35 AM Revision 11ff6569 (git): enumerator.c: Fix ArithmeticSequence for complex step
- Make sure Enumerator::ArithmeticSequence#each to work well
for a complex step value.
This reverts commit ca47fb329a1d48af3e1009620bdb18e931c9f188.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64695 b2dd03c8-39d4-4d8f-98ff-823fe69b... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64694 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 05:52 AM Revision ca47fb32 (git): Revert "enumerator.c: Fix ArithmeticSequence for complex step"
- This reverts commit 0a0f8238d02a2dfff4cd2892408e14cb826cec7e.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64693 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
05:44 AM Feature #14136: Implement #empty? on more classes
- bozhidar (Bozhidar Batsov) wrote:
> > Yeah, in this case I'd argue that it's better to use some `top-level` methods of `File` instead, but in general every object that has the notion of size should also have the option of emptiness. Tha... -
05:12 AM Feature #14136: Implement #empty? on more classes
- bozhidar (Bozhidar Batsov) wrote:
> phluid61 (Matthew Kerwin) wrote:
> ...
Also I'm curious who'd claim that adding an empty method doesn't make sense for something like `Tempfile` or `StringIO`. -
- phluid61 (Matthew Kerwin) wrote:
> mikegee (Michael Gee) wrote:
> ...
Yeah, in this case I'd argue that it's better to use some `top-level` methods of `File` instead, but in general every object that has the notion of size should also ... - 05:35 AM Revision 0a0f8238 (git): enumerator.c: Fix ArithmeticSequence for complex step
- Make sure Enumerator::ArithmeticSequence#each to work well
for a complex step value.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64692 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
04:04 AM Revision 0f6708eb (git): resurrect the string to expect modifications
- String#freeze can be redefined to be destructive. While such
redefinition is definitely weird, it should be possible. Resurrect
the string to prepare for that sort of things.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64691 b2... -
- Simply use DISPATCH_ORIGINAL_INSN instead of rb_funcall. This is,
when possible, overall performant because method dispatch results are
cached inside of CALL_CACHE. Should also be good for JIT.
----
trunk: ruby 2.6.0dev (2018-09-12 t... -
-
- This instruction can be written without rb_funcall. It not only boosts
performance of case statements, but also makes room of future JIT
improvements. Because opt_case_dispatch is about optimization this
should not be a bad thing to hav...
09/11/2018
-
11:01 PM Feature #15074: Create 'official' C API documentation on ruby-doc.org
- @hsbt what do you think about writing the docs within C API files like `ruby.h`, `intern.h` and `extension.h` itself so that ruby-doc.org can run doxygen on them to generate the relevant documents? It will keep the documentation up-to-da...
-
- Alright. I think its a matter of adding the extension.rdoc file to the ruby-doc.org sources and then push changes to extension.rdoc in the core ruby repo.
-
09:55 PM Bug #15105 (Closed): `rb_debug_inspector_open` breaks lazy proc optimization
- Calling `rb_debug_inspector_open` inside a block that uses lazy proc optimizations breaks the implementation. r60397 introduced the bug.
I've attached a failing test, but I'll paste the diff here as well:
```diff
From 54e55b83f1a... -
04:16 PM Bug #15103: undecipherable nil error for `+`
- Never mind, I found the error. Please close this.
They were not operationally exact.
This ``while num >= modk + residues[r]; r += 1 end``
had to be this ``while (num-1) >= modk + residues[r]; r += 1 end``
Now I remember why I did... -
- I am experiencing the following error in a method I rewrote that makes no sense whatsover.
Here's the original method.
```
def method1(num, residues, flag)
modpg, rescnt = (residues[-1] - 1), residues.size
... - 03:14 PM Revision bddc28b2 (git): * 2018-09-12
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64688 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- if and only if it's not moved yet, to avoid potential bugs in the future.
_mjit_compile_send.erb: ditto
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64687 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
03:12 PM Bug #15082: Memory leak in net/http/response and net/http/header
- Thanks for pointing this PR. I ran it again with this specific memory_profiler version and there is no memory leaks.
It seems that I don't have enough permissions to close the issue. -
05:16 AM Bug #15082: Memory leak in net/http/response and net/http/header
- @alexis Look into this PR https://github.com/SamSaffron/memory_profiler/pull/59
Once merged this PR should fix the issue you faced.
I think this can be closed now. -
- To attempt to fix CI failure on rubyci freebsd:
https://rubyci.org/logs/rubyci.s3.amazonaws.com/freebsd11zfs/ruby-trunk/log/20180911T123001Z.fail.html.gz
```
JIT success (68.7ms): mjit9@(eval):1 -> /usr/home/hsbt/chkbuild/tmp/build/2018... -
- to raise descriptive KeyError instead of NoMethodError in case these
attrs are accidentally removed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64685 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- rather than `trace_enablable_insns` which is hard to maintain.
This would make performance bad because new branches for tracepoint will
be added.
Optcarrot before:
fps: 56.885371547337655
fps: 60.02493636060194
fps: 63.051028327122076
... -
- when catch_except_p is false and insn.always_leaf? is true (never makes
arbitrary method call in the insn).
On Optcarrot, unfortunately this didn't have measureable performance impact.
But still this is a good direction since it becomes... -
- as well, to make CI succeed with VM_CHECK_MODE > 1.
vm_insnhelper.c: drop unnecessary MJIT_HEADER ifdef. This is intended to
be ignored by having `static inline`. Removing that by macro would be
helpful for minimizing compilation time, ... -
- failure
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64681 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64680 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64679 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64678 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- An instruction is leaf if it has no rb_funcall inside. In order to
check this property, we introduce stack canary which is a random
number collected at runtime. Stack top is always filled with this
number and checked for stack smashing... - 09:46 AM Revision ecda2d23 (git): * 2018-09-11
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64676 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64675 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
01:18 AM Feature #15092 (Rejected): Provide step count in Range constructor
-
- OK I'm convinced this is probably a bad idea since `step` returns an `ArithmeticSequence`
and is not a property of `Range`.
Closing the issue. -
- We've already have Emuerator::ArithmeticSequence in trunk.
Please use it. -
12:10 AM Bug #13931: correct install_name of libruby on macOS (libruby.2.5.0.dylib -> libruby.2.5.dylib)
- Requesting backport to 2.4
https://github.com/Homebrew/homebrew-core/pull/31967#issuecomment-420100406
09/10/2018
-
11:21 PM Misc #14981: DevelopersMeeting20180913Japan
- * [Feature #13167] Dir.glob is 25x slower since Ruby 2.2 (ahorek)
* improves performance of a glob pattern with braces, especially on windows but also on other platforms
* the idea was already approved by @matz #13873 but later rev... -
07:59 PM Bug #15088: Leaky behaviour when GC disabled
- Thank you @normalperson & @ko1 for taking a look and for the explanation here 😀
-
07:18 PM Bug #15097: Gem install fails on Ruby 2.5.1 with Cygwin (get_dns_server_list undefined)
- nobu (Nobuyoshi Nakada) wrote:
> I can't reproduce it, with the trunk and 2.5.2 in the repository, using
Thanks for taking a look at it.
This is on a 32-bit Windows, if that has any significance?
However I now figured out a work... -
- I can't reproduce it, with the trunk and 2.5.2 in the repository, using
```
$ LANG=C i686-pc-cygwin-gcc -v
Using built-in specs.
COLLECT_GCC=i686-pc-cygwin-gcc
COLLECT_LTO_WRAPPER=/usr/lib/gcc/i686-pc-cygwin/6.4.0/lto-wrapper.exe
... -
- Compiled and installed Ruby 2.5.1 on Cygwin, and now gem install fails with:
~~~ ruby
ERROR: While executing gem ... (NameError)
undefined local variable or method `get_dns_server_list' for Win32::Resolv:Module
~~~
While ... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64674 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64673 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
07:23 AM Feature #15074: Create 'official' C API documentation on ruby-doc.org
- At first, `ruby-doc.org` is not our `official` web resource. It's a third party resource.
But `ruby-doc.org` maybe uses documentation generated by rdoc with ruby source code. So, You can add its documentation under the `doc` directory. -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64672 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * util.c (ruby_qsort): __STDC_VERSION__ may not be defined even if it
is available. fixed duplicate definitions when qsort_s is available
on non-Windows platforms. [ruby-core:88921] [Bug #15091]
git-svn-id: svn+ssh://ci.ruby-lang.... - 01:43 AM Revision 5679d0c5 (git): there is no such predefined macro. check build before commit!
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64670 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
09/09/2018
-
- Koichi Sasada <ko1@atdot.net> wrote:
> On 2018/09/10 2:26, Eric Wong wrote:
> > Sorry, I mean they can share VM stack cache; not the in-use VM stack.
>
> It makes sense.
>
> But I remember that Fiber stack is smaller than Thread... -
06:03 PM Feature #15095: [PATCH] share VM stack between threads and fibers if identical
- On 2018/09/10 2:26, Eric Wong wrote:
> Sorry, I mean they can share VM stack cache; not the in-use VM stack.
It makes sense.
But I remember that Fiber stack is smaller than Thread VM stack because
of rough estimation (Fiber may... -
- On 2018/09/10 2:31, Eric Wong wrote:
> OK, I thought that was the case. But no plans to use `const' here
> in the future, right? In other words, this patch should be OK?
Not sure about the future.
But it is okay to commit your pat... -
- ko1@atdot.net wrote:
> normalperson (Eric Wong) wrote:
> > Also, why cast in ec_set_vm_stack?
> > https://80x24.org/spew/20180909080140.16293-1-e@80x24.org/raw
> > I've barely touched code this week, so maybe I forgot something :x... -
- ko1@atdot.net wrote:
> > ec->vm_stack is always allocated with malloc, so stacks for root
> > fiber (thread stack) and non-root fibers can be shared as long
> > as the size is the same.
>
> I'm not sure why we can share the VM sta... -
- normalperson (Eric Wong) wrote:
> Also, why cast in ec_set_vm_stack?
> ...
Yes. These fields were `const` at least on my laptop to modify something.
-
- > ec->vm_stack is always allocated with malloc, so stacks for root
> ...
I'm not sure why we can share the VM stack. If we run root fiber and non-root fiber with switching, it will conflicts.
Do I misunderstand something?
-
- Also, why cast in ec_set_vm_stack?
https://80x24.org/spew/20180909080140.16293-1-e@80x24.org/raw
I've barely touched code this week, so maybe I forgot something :x -
- ec->vm_stack is always allocated with malloc, so stacks for root
fiber (thread stack) and non-root fibers can be shared as long
as the size is the same. The purpose of this change was mainly
to reduce dependencies on ROOT_FIBER_CONTE... - 03:46 PM Revision 73e7cdeb (git): * 2018-09-10
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64669 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64668 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Thank you Eric.
This is my 2nd experiment:
```
...
# adding the following code to original trial code.
GC.start
puts"After2 - All: #{ObjectSpace.each_object.count}"
puts"After2 - Symbols: #{Symbol.all_symbols.size}"
```
... -
02:01 PM Feature #15093: Retrive Array elements with infinite ranges specified by Float::INFINITY
- This could be fixed, at least in theory, but in practice this is no longer an issue: use endless range
```
a[0...] # => [1, 2, 3, 4]
```
I'm unsure if we should change the behavior, since it allows catching of subtle errors where... -
09:24 AM Feature #15093: Retrive Array elements with infinite ranges specified by Float::INFINITY
- I can not say much about the reasoning (I am neutral there) but
I agree in one regards - if we use a large number, we get all
Array members:
a = [1,2,3,4] # => [1, 2, 3, 4]
a[0..111111111] # => [1, 2, 3, 4]
So from thi... -
- Currently in Ruby head (2.6) the following gives an error:
~~~ ruby
2.6-head :001 > a = [1,2,3,4]
# => [1, 2, 3, 4]
2.6-head :002 > a[0..Float::INFINITY]
#Traceback (most recent call last):
# 2: from /home/sameer/.rvm/rubie... -
- v0dro (Sameer Deshmukh) wrote:
> Advantages of changing the constructor:
> ...
Very debatable.
> * Consistency in specifying step count in constructor (simple and straightforward way) and using
> ...
It's actually not consistent, s... -
- > Wouldn't the result be [1, 4, 7, 10, 13]?
Off-by-one is ... common. :)
A bit more on topic, Float::INFINITY is quite long. Could
we not use :infinity to refer to it in some methods or
something shorter? I think ruby users shou... -
- v0dro (Sameer Deshmukh) wrote:
> I would like to propose making changes to the Range constructor so that a user can specify
> ...
The array will return `[1,4,7,10]`. Sorry for previous mistake. -
- > Wouldn't the result be [1, 4, 7, 10, 13]?
Ah yes. My bad. Editing the description. Thank you. -
03:12 AM Feature #15092: Provide step count in Range constructor
- v0dro (Sameer Deshmukh) wrote:
> And I want the values `1`, `4`, `7`, `10` and `12`, I can simply specify a Range like this:
> ...
Wouldn't the result be `[1, 4, 7, 10, 13]`?
-
- Advantages of changing the constructor:
* Makes it easy to read test code for someone new to Ruby.
* Consistency in specifying step count in constructor (simple and straightforward way) and using
the shorter syntax (idiomatic Ruby way... -
- No need to change the constructor. Instead of:
```
Range.new 0, Float::INFINITY, 3
```
Use the shorter
```
(0...) % 3
```
-
I would like to propose making changes to the Range constructor so that a user can specify
a `step` count along with `start` and `stop`. Since Ruby 2.6 will introduce a `step` property
in Ranges anyway I think this will be a useful a...-
- thanks @nobu
but now I have an error
```
error Unknown version qsort_s
```
the build passed without this line
https://github.com/ruby/ruby/commit/8e9a8c043ef5b16a2f13999bd2c5d59d61dd00d1#diff-6d7e72ea1ca2b7d0258e568a155084bdR219 -
- Applied in changeset trunk|r64661.
----------
util.c: qsort_s in C11
* configure.ac: macro for C11 to use qsort_s.
* util.c (ruby_qsort): fix for C11 qsort_s. the comparison function
for MSVCRT qsort_s is compatible with BSD qsort_... -
- [Fix GH-1937]
Co-authored-by: Jun Aruga <jaruga@redhat.com>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64667 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64666 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64665 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64664 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Koichi Sasada <ko1@atdot.net> wrote:
> Ok. We need to avoid this kind of crash.
>
> https://bugs.ruby-lang.org/issues/15050
OK, I think I fixed the issue (also including [Feature #15095])
Greg, there's a small win32-specific c... - 08:50 AM Revision 64b326c2 (git): spec/ruby/library/socket/addrinfo: require for SocketSpecs
- Otherwise, I get NameError when running these IPv6 tests individually
or in parallel.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64663 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 07:49 AM Revision 432af3e6 (git): * 2018-09-09
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64662 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * configure.ac: macro for C11 to use qsort_s.
* util.c (ruby_qsort): fix for C11 qsort_s. the comparison function
for MSVCRT qsort_s is compatible with BSD qsort_r, but not with C11
qsort_s, in spite of its name.
note that mingw ... -
- nobu (Nobuyoshi Nakada) wrote:
> Doesn't the order matter?
Very often, it does not. If it does, one can always require the one that's needed first, say, then require the whole directory; require won't load the same file twice so this... -
- shevegen (Robert A. Heiler) wrote:
> In the event that the project may be relocated, here is the
> ...
Doesn't the order matter? -
- After working a lot on `**{}`, I still strongly believe that we must maintain conversion of keyword arguments to positional argument, e.g.:
```
def foo(*ary)
end
foo(kw: 1) # => must remain ok
```
OTOH, it may be possible to di... -
- mame (Yusuke Endoh) wrote:
> I'd like to agree with you, but also really like to fix #15052. Do you find a good semantics of **hash to satisfy both this ticket and that one?
This is a long post, sorry.
* Summary *
I agree that ... -
- Alright. So as a first step can you please tell me how it would be possible to create
this page for the C API? Or maybe start with the wiki?
I can send a PR with some preliminary docs once you approve a course of action.
09/08/2018
-
- pdahorek@seznam.cz wrote:
> https://bugs.ruby-lang.org/issues/15091
> ```
> https://www.oracle.com/technetwork/server-storage/solaris11/downloads/index.html
Weird, I'm not familiar with Solaris versions, but it seems our
CI machi... -
- gcc -v
```
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/usr/gcc/7/lib/gcc/x86_64-pc-solaris2.11/7.3.0/lto-wrapper
Target: x86_64-pc-solaris2.11
Configured with: /builds/ulhg/workspace/Solaris_11u4/Userland/full-build/... -
- I thought about creating a new issue but then I remembered that
the issue here refers to a similar use case that I wanted to
show.
Take the following link as an example:
https://github.com/jordansissel/fpm/blob/master/lib/fpm.rb
... -
- As long as I can see from the benchmark result for the improved case, it looks good. But at least I would like to see micro benchmarks for opt_send_without_block and send. Because of _mjit_compile_send, it may not be affected so much tho...
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64660 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 01:09 AM Revision edc4df32 (git): * 2018-09-08
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64659 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 01:09 AM Revision e83c3083 (git): test/ruby/test_thread.rb: join threads in each test
- Leaky thread detection doesn't happen frequently, enough
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64658 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
09/07/2018
-
- liamtabi@gmail.com wrote:
> https://bugs.ruby-lang.org/issues/15088
> 100_000.times do |i|
> "to_sym_#{i}".to_sym
> end
>
> GC.enable
> GC.start
You need at least another call to GC.start, here.
I think it is because g... -
- Hello
I've observed some leaky behaviour when creating dynamic Symbols ("string".to_sym) when GC is disabled. Creating a dynamic symbol after running `GC.disable` creates an extra string allocation that isn't garbage collected when GC... -
04:36 PM Bug #15078: Hash splat of empty hash should not create a positional argument.
- marcandre (Marc-Andre Lafortune) wrote:
> First, I hope we can agree that `any(**{})` and `any(**Hash.new)` should have the exact same result (for any definition of `any`).
Of course :-)
> ...
What do you think about #15052?
... -
- Good catch. My patch for #15052 will fix this issue.
-
02:58 PM Bug #15087 (Closed): Segmentation fault with splat and block
- The following code produces segmentation faults in Ruby 2.5.0, 2.5.1, and 2.6.0dev (2018-09-07 trunk 64656). This code worked in Ruby 2.4.1.
~~~ ruby
test_proc = Proc.new { "test" }
def foo(bar: 1, &block)
yield
end
foo(**{}, &... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64657 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
>
> ...
As Benoit says (rightly), I don't think we need a very elaborate API. Therefore the following functions would be sufficient:
* Constructor from double: Something like `rb_complex_new(double real, double imag)`.
* Access r...-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64656 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * ext/objspace/object_tracing.c (trace_object_allocations_start): to
prevent TracePoint objects from GC, register them in the VM, since
they are unique per VM.
http://ci.rvm.jp/results/trunk-test@ruby-sky3/1291901
* ext/objspace/o... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64654 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- [Fix GH-1946]
From: Josh Cheek <josh.cheek@gmail.com>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64653 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64652 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 03:16 AM Revision 870c296d (git): * 2018-09-07
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64651 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64650 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
09/06/2018
-
11:25 PM Feature #15085 (Rejected): Decrease memory cache usage of MJIT
- MJIT makes ruby-methods faster by ordinary, but I have observed that some cases are exceptional.
I guess the one is caused by `invokesuper` instruction.
And I guess it is related to memory caching, especially iTLB.
Attached "export-... -
- First, I hope we can agree that `any(**{})` and `any(**Hash.new)` should have the exact same result (for any definition of `any`).
So the question is: what should the result be?
For me, writing `**hash` means "do as if I had writte... -
03:51 AM Bug #15078: Hash splat of empty hash should not create a positional argument.
- It is also related to
https://bugs.ruby-lang.org/issues/11860
https://bugs.ruby-lang.org/issues/12022
https://bugs.ruby-lang.org/issues/10708 -
05:44 PM Bug #14716: SecureRandom throwing an error in Ruby 2.5.1
- I worked around this problem by forcing `SecureRandom` to use the OpenSSL gem:
```ruby
module SecureRandom
if RUBY_VERSION == "2.5.1"
class << self
def gen_random(n)
begin
require 'openssl'
rescue N... -
- lars@greiz-reinsdorf.de wrote:
> Issue #14999 has been updated by larskanis (Lars Kanis).
> > > So it seems it doesn't happen in Ruby <= 2.5 due to thread
> > > caching being disabled then?
> >
> > Right; but that might just ... -
- This affects 2.4.4 as reported by Lars: [ruby-core:88852]
I expect it to affect all 2.x versions, even. [Bug #14999]
"thread_sync.c (rb_mutex_sleep): disable interrupt checking in ensure"
Should be straightforward despite the ti... -
- Upon further inspection, it seems duplicating long text with interpolation allocates an extra string(frozen). You can observe this using the following script.
~~~
def string_info
GC.disable
obj_ids = []
final_ids = []
obj_s... -
- It can be an issue related to MemoryProfiler or some edge case to ObjectSpace.trace_object_allocations_start
The following code also result in a similar result:
~~~
MemoryProfiler.report{ 200.times{|i| "SOME RANDOM TEXT#{i}: SOME ... -
- Hello,
I'm observing a memory leak in net/http with the following script :
~~~
require "net/http"
require "bundler/inline"
gemfile do
gem "memory_profiler"
end
def profile_http_get(n)
uri = URI("http://www.ruby-lang.... - 09:24 AM Revision ad10b43f (git): re-patched r50599 because of test failure on TravisCI.
- [Backport #14060]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_4@64649 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- This issue has been masked by #15072, but has resurfaced.
@nobu: Ping! -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64648 b2dd03c8-39d4-4d8f-98ff-823fe69b080e