Activity
From 07/01/2018 to 07/07/2018
07/07/2018
- 11:59 PM Revision 4719a454 (git): * 2018-07-08
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63878 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 11:59 PM Revision ac41c2c1 (git): process.c (rb_f_exec): pause MJIT before replacing process
- Non-parallel "make test-spec" caused
spec/ruby/core/process/wait2_spec.rb failures because mspec
uses "exec" in single-process mode, so there's no chance
the post-exec state could know about the MJIT child process
from its pre-exec state... -
10:52 PM Bug #14867: Process.wait can wait for MJIT compiler process
- takashikkbn@gmail.com wrote:
> Hi Eric. I fixed [Bug #14892] on r63875, but only the rubyspec check you added is failing with -DMJIT_FORCE_ENABLE like:
>
> ```
> $ make test-spec
Thanks, I think it's a bug of rb_f_exec and it se... -
02:58 PM Bug #14867 (Assigned): Process.wait can wait for MJIT compiler process
- Hi Eric. I fixed [Bug #14892] on r63875, but only the rubyspec check you added is failing with -DMJIT_FORCE_ENABLE like:
```
$ make test-spec
$ /home/k0kubun/src/github.com/ruby/ruby-svn/.ruby-force/miniruby -I/home/k0kubun/src/gith... -
08:48 PM Bug #14900: Extra allocation in String#byteslice
- Nice catch I will try to verify on my end too
-
09:47 AM Bug #14900 (Closed): Extra allocation in String#byteslice
- When executing `String#byteslice` with a range, I noticed that sometimes the original string is allocated again. When I run the following script:
~~~ ruby
require "objspace"
string = "a" * 100_000
GC.start
GC.disable
generati... -
05:43 PM Feature #13620: Simplifying MRI's build system: always make install
- normalperson (Eric Wong) wrote:
> Of course, there will be differences not detected before
> ...
There are many differences, and fundamental things like RbConfig.ruby pointing to the current ruby are broken (see the original descripti... -
- The patch got outdated in the latest trunk, so I updated it again. It's a really small change with big impact on memory usage when uploading files.
-
- * [Bug #14878] Add command line argument to deactivate JIT (k0kubun)
* Please discuss the necessity of the flag and its name in the proposal. -
- This is not working with cppflags="-DMJIT_FORCE_ENABLE" on my machine.
```
$ make test-spec
$ /home/k0kubun/src/github.com/ruby/ruby-svn/.ruby-force/miniruby -I/home/k0kubun/src/github.com/ruby/ruby-svn/lib /home/k0kubun/src/github.com/... -
- Applied in changeset trunk|r63875.
----------
insns.def: stop pushing unnecessary keys for MJIT
[Bug #14892] -
- [Bug #14892]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63875 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- for catch_except_p case
Partially solving Bug#14892
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63874 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
08:45 AM Bug #14899 (Closed): Pathname#glob Matching to the directory results in lack
- Pathname.glob では問題ないのですが、
ruby 2.5 から追加された Pathname#glob で、スラッシュで終わる引数を渡すとディレクトリ名の最初の１文字がかけた状態で返ってきます
~~~
Pathname('/tmp/test').glob('**/*') => [#<Pathname:/tmp/test/abc>, #<Pathname:/tmp/test/123>]
Pathname.glob('/tmp/test/**/*/... -
05:52 AM Bug #14898: test/lib/test/unit/parallel.rb: TestSocket#test_timestamp stuck sometimes
- On 2018/07/07 14:36, Eric Wong wrote:
> Maybe r63872 can help by retrying send.
Great! Thank you.
--
// SASADA Koichi at atdot dot net -
- Koichi Sasada <ko1@atdot.net> wrote:
> On 2018/07/06 18:47, Eric Wong wrote:
> > I would not expect it to be (but am not sure), I think it's
> > possible the kernel could drop packets if under memory pressure.
>
> mmm. can we rewr... -
- Applied in changeset trunk|r63872.
----------
test/socket/test_socket.rb (test_timestamp): retry send
I theorize there can be UDP packet loss even over loopback if
the kernel is under memory pressure. Retry sending periodically
until ... -
- On 2018/07/06 18:47, Eric Wong wrote:
> I've never seen it stuck myself.
Only a few times per thousands trial. I also never seen in manual trial.
> Is UDP over loopback supposed to be reliable?
Maybe yes because other tests pa... - 05:34 AM Revision 886cfe49 (git): * 2018-07-07
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63873 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 05:34 AM Revision eb78beda (git): test/socket/test_socket.rb (test_timestamp): retry send
- I theorize there can be UDP packet loss even over loopback if
the kernel is under memory pressure. Retry sending periodically
until recvmsg succeeds.
i[ruby-core:87842] [Bug #14898]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63...
07/06/2018
-
- > https://bugs.ruby-lang.org/issues/14851
Maybe reject this feature. Keeping non-sleepy timer thread
implementation (I will call it "polling timer thread") since
manually enabling it can find bugs with thread_win32 platform. -
- funny.falcon@gmail.com wrote:
> It is impossible to implement Thread migration between native
> threads. All other is possible.
It may be possible but we'd either lose performance and/or break
C extensions. So it won't happen soo... -
06:10 PM Feature #13618: [PATCH] auto fiber schedule for rb_wait_for_single_fd and rb_waitpid
- > Yes, they are great, but it's probably impossible to implement in Ruby.
It is impossible to implement Thread migration between native threads. All other is possible.
> ...
Don't get me wrong: most of code is linear and thread-saf... -
- > But we have example of Go: goroutine is a really green thread, and they do really scale.
Yes, they are great, but it's probably impossible to implement in Ruby, and it still requires a lot of non-trivial synchronisation.
> ...
Ye... -
- > In my experience, threads simply don't scale as as well as fibers, there is too much overhead.
Native threads doesn't scale.
But we have example of Go: goroutine is a really green thread, and they do really scale.
And Ruby 1.8 ha... -
01:57 PM Feature #14328 (Closed): SIMD vectorization
-
- * variable.c (rb_const_search): call #const_missing method on
private constants, as well as uninitialized constants.
[Feature #14328]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63871 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- ko1@atdot.net wrote:
> With parallel tests (`make test-all TESTS=-j4` with 4 parallelism) stuck sometimes.
>
> http://ci.rvm.jp/results/trunk-test@ruby-sky3/1087178
>
> We can see this stuck very old revisions but not sure how to... -
- With parallel tests (`make test-all TESTS=-j4` with 4 parallelism) stuck sometimes.
http://ci.rvm.jp/results/trunk-test@ruby-sky3/1087178
We can see this stuck very old revisions but not sure how to solve...
Anyone help us?
-
- * compile.c (iseq_peephole_optimize): remove unreachable jump
instruction only. if it is labeled and referred from other
instructions, it is reachable and must not be removed.
[ruby-core:87830] [Bug #14897]
git-svn-id: svn+ssh://... -
- Greg.mpls@gmail.com wrote:
> Well done. The build is 'green' (passing).
Thanks! Committed as r63869. -
- Applied in changeset trunk|r63869.
----------
mjit.c: fix waitpid macro return value for win32
We started checking return value of waitpid, so it needs
to be correct for win32 platforms for MJIT to work.
Thanks-to: MSP-Greg (Greg L) <... -
04:38 AM Bug #14867: Process.wait can wait for MJIT compiler process
- Eric,
Well done. The build is 'green' (passing). Script generated result summary is:
```
——————————————————————————————————————————————————————————————————————————————— Test Results
0 Total Failures/Errors ... -
- Greg: oh, I think I finally spotted it. The waitpid() macro for
_WIN32 doesn't set a return value. The old code assumed waitpid
always succeeds, and the new code checks the return value.
So I think we need to fake the return valu... -
- @normalperson
> Were there any compiler warnings building process.c or mjit.c?
No. Only warning were for `warning: flip-flop is deprecated` in another file...
> ...
Yes, but it's a mess with STDOUT having the line(s) added by t... -
- Greg.mpls@gmail.com wrote:
> Below is the out from `ruby --jit -e exit`
>
> ```
> C:\Greg\GitHub> ruby --jit -e exit
> process.c: sigchld=0 getpid.size=8 rb_pid_t.size=8
> mjit.c: sigchld=0 getpid.size=8 pid_t.size=8
> mjit.c: _... -
- @normalperson
Attached is the config portion of ruby-loco mingw build log.
Thanks, Greg -
- @normalperson
Below is the out from `ruby --jit -e exit`
```
C:\Greg\GitHub> ruby --jit -e exit
process.c: sigchld=0 getpid.size=8 rb_pid_t.size=8
mjit.c: sigchld=0 getpid.size=8 pid_t.size=8
mjit.c: _WIN32 __MINGW32__ __MINGW6... -
- @normalperson
I've got builds running both locally and on Appveyor, when they finish I'll have more info.
> Just to reiterate, "make test" (not "test-all") passes?
Actually, btest, test-basic, & test-spec all pass.
> ...
If t... - 06:50 AM Revision 79f01d39 (git): mjit.c: fix waitpid macro return value for win32
- We started checking return value of waitpid, so it needs
to be correct for win32 platforms for MJIT to work.
Thanks-to: MSP-Greg (Greg L) <Greg.mpls@gmail.com>
[ruby-core:87832] [Bug #14867]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby... -
04:53 AM Bug #14897: Unexpected behavior of `if` in specific code
- Good catch. It was a bug of peephole optimization. Fixed. Thank you.
-
- Applied in changeset trunk|r63868.
----------
Fix a bug of peephole optimization
```
if L1
L0:
jump L2
L1:
...
L2:
```
was wrongly optimized to:
```
unless L2
L0:
L1:
...
L2:
```
To make it conservative, this optimizatio... -
04:03 AM Bug #14897: Unexpected behavior of `if` in specific code
- Seems `obj = obj` is (mistakenly) optimized out.
```
% ./miniruby --dump=i -v tmp.rb
ruby 2.6.0dev (2018-07-06 trunk 63854) [x86_64-darwin15]
== disasm: #<ISeq:<main>@tmp.rb:1 (1,0)-(10,16)> (catch: FALSE)
0000 putspecialobject ... -
03:22 AM Bug #14897 (Closed): Unexpected behavior of `if` in specific code
- I found a strange behavior of `if` in the following code.
~~~ ruby
def seems_bug(obj)
if obj || obj
obj = obj
else
raise obj.inspect
end
obj
end
seems_bug('foo')
#=> RuntimeError: "foo"
~~~
This code is... -
- ```
if L1
L0:
jump L2
L1:
...
L2:
```
was wrongly optimized to:
```
unless L2
L0:
L1:
...
L2:
```
To make it conservative, this optimization is now disabled when there is
any label between `if` and `jump` instructions.
Fix... - 02:15 AM Revision fcf8c622 (git): thread.c (ruby_ppoll): improve timespec to msec conversion
- Round up non-zero <1ms timeouts to 1ms and use INT_MAX instead
of infinite (-1) for extremely large timeouts. All of our
ppoll/select callers are able to handle spurious wakeups,
anyways.
This avoids excessive CPU usage and busy waits ... - 02:15 AM Revision 24aa7a13 (git): thread.c: our fake ppoll implementation is static
- Rename it to "ruby_ppoll" so it looks more obvious in debuggers.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63866 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
01:53 AM Revision 043d22cb (git): Fixup r63864
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63865 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63864 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
12:49 AM Bug #14892: r63763 breaks -DMJIT_FORCE_ENABLE=1
- Many cases (including `a, b = nil`) go well after r63863.
Now I found that there remain 2 issues at least.
1. `pc` is set to `next_pos` when insn is `opt_xxx` and `body->catch_except_p` is falsy.
* ex. `1000.times { break if /a/ ...
07/05/2018
-
- > They have their own IO scheduler because ruby had just native threads, which are bad as IO scheduler.
Thanks so much for your answer, it's very detailed and gives me a clear picture about what you are thinking.
In my experience, ... -
- > It's also going to be more useful for existing code bases like ActionCable, Puma, Async, etc which use their own IO scheduler.
They have their own IO scheduler because ruby had just native threads, which are bad as IO scheduler.
Ok... -
- Thanks for your feedback.
> About "extendable api" vs "implicit behavior" (ie between your proposal
> ...
I don't see how that comparison applies to `Thread.scheduler` PR. Because it's completely transparent to higher level code.
... -
- @ioquatix About "extendable api" vs "implicit behavior" (ie between your proposal
for "Thread.scheduler=' and replacing "Thread" with green implementation),
I'd prefer "implicit behavior".
Because, there were already EM::Synchrony, th... -
- Looks like I was not exactly right: typical iteration over Enumerator doesn't use Fiber.yield.
But usage of Enumerator as external iterator does. And `zip` method does use external iterator.
```
> def aga; yield 1; Fiber.yield 4; yiel... -
- Greg.mpls@gmail.com wrote:
> @normalperson
>
> Thanks for the patch, I added it, and a build with r63859 locked/froze on test-all. I’m going to see if I can get any more info by building/testing locally. I’m surprised and wonderin... -
- @normalperson
Thanks for the patch, I added it, and a build with r63859 locked/froze on test-all. I’m going to see if I can get any more info by building/testing locally. I’m surprised and wondering if something reverted wasn’t re-a... -
- Greg.mpls@gmail.com wrote:
>
> Importantly, mswin (ruby's vc12) passed r63856.
Good to know. I saw your message before
http://mswinci.japaneast.cloudapp.azure.com/vc12-x64/ruby-trunk/recent.html
got updated, I guess there is a ... -
- Importantly, mswin (ruby's vc12) passed r63856.
Unfortunately, ruby-loco locked up on test-all. So, before the reversion (r63852), test-all completed, but all JIT tests failed. After the reversion, all JIT tests passed.
Ok, I'l... -
- takashikkbn@gmail.com wrote:
> As it was reverted on r63852 by @naruse, assigning this back
> to Eric. I'm guessing it was reverted since it was not
> portable mainly for Windows. Whenever you commit things, you
> can check at least... -
- Greg.mpls@gmail.com wrote:
> I'm not sure what you look at on the web (GitHub, Appveyor), so briefly:
> 1. The mswin builds do not test MJIT, but mingw (ruby-loco)
> does. Note that mingw tests 3 time a day, so I may not know
> w... -
- @normalperson
I'm not sure what you look at on the web (GitHub, Appveyor), so briefly:
1. The mswin builds do not test MJIT, but mingw (ruby-loco) does. Note that mingw tests 3 time a day, so I may not know what revision caused wh... -
09:10 PM Bug #14895: Inconsistent constant names when using const_set on a singleton class
- This looks like a bug to me.
Here's a simplified example:
```
s = Object.new.singleton_class
a = s.const_set(:A, Module.new) # => #<Module:0x00007fed619915f0>
b = s.class_eval "module B; self end" # => #<Class:0x00007fed6... -
05:35 PM Bug #14895: Inconsistent constant names when using const_set on a singleton class
- So, what you're saying is that the singleton class is being assigned to a constant when it creates module `A`, so modules are only named after this step?
Is this expected behavior then? -
- silver_phoenix (Pedro Pinto) wrote:
> Yes, I understand that, but why isn't `Z` named like `B` is?
> ...
Because they are not "fully-named" yet.
-
- Yes, I understand that, but why isn't `Z` named like `B` is?
`Z` is `#<Module:...>` (unnamed)
`B` is `#<...>::B` (named)
There's something not being initialized on the singleton class, which is only triggered after adding a named ... -
- The names of `A`, `B`, and `Z` are temporary names at the moments of assignment.
Once you assign `x.singleton_class` to a constant, say `C`, they appear as `C::A`, `C::B`, and `C::Z`, respectively. -
06:32 PM Feature #14887: Array#delete_if does not use #delete
- marcandre (Marc-Andre Lafortune) wrote:
> You probably meant "... call #delete_at in the background".
Yes, Marc-André, that's what i want to mean...
> ...
I don't known ruby background. If it seems impossible, forget about that. -
07:28 AM Feature #14887: Array#delete_if does not use #delete
- @shevegen : the problem is that:
* `delete_if` can't call `delete` for the given reason of duplicates. the example with `with_index` was only an example to show where it differ.
* `delete_if` can't call `delete_at` because it would cau... -
07:05 AM Feature #14887: Array#delete_if does not use #delete
- This is actually a (to me) quite interesting discussion. I want to comment on
a very few things but of course I'll keep in mind that there is a specific
suggestion in the issue which should not be forgotten.
Sylvain suggested, essen... -
- r63655 was tightly coupled to handle_frames and some assumptions seems
to have been broken by r63763.
To partially resolve Bug#14892, this reverts the optimization for now. I
want to make MJIT CI happy first and then I'll probably retry... - 03:20 PM Revision 96f1c663 (git): * 2018-07-06
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63862 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63861 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
12:48 PM Revision 1d74de37 (git): ruby tool/update-deps --fix
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63860 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63859 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63858 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * Both timegmw and gmtimew ignores leap second if the timezone doesn't
have leap seconds on the first call of init_leap_second_info()
* Add Bug::Time.reset_leap_second_info for testing
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk... -
- duerst (Martin Dürst) wrote:
> On the other hand, if we change it to the source encoding, then the implementation gets more difficult (it has to somehow get the source encoding). Also, some programs that use this functionality in a setu... -
- Yes, I agree with what you say, and I agree with your conclusion, I was just giving an example where it failed with async which highlights the issue :)
-
- I've shown `to_enum(:aga).to_a` to present the place where I wasn't right.
But if you look at your own second example, you will see that it doesn't do what it should do
(if `Fiber.yield` is replaced with yield point of your scheduler... -
- For the first case, you naturally can't call Fiber.yield in that context... but this works:
```
#!/usr/bin/env ruby
require 'fiber'
fiber = Fiber.new do
def aga; yield 1; Fiber.yield 4; yield 8; end
puts to_enum(:aga).to_a
... -
- Yes. While usually Enumerator doesn't call to Fiber.yield, it is called if Enumerator is used as external iterator:
```
> def aga; yield 1; Fiber.yield 4; yield 8; end
> ...
Traceback (most recent call last):
6: from /usr/bin... -
- Are you saying that calling Fiber.yield is not valid within an enumerator?
-
07:36 AM Feature #14697 (Assigned): Introducing Range#% as an alias to Range#step
-
- Here is what I made using RubyVM::AST. It was useful. https://github.com/ioquatix/covered
In the end I made a regular expression to match node type. It was better than I expected. -
- And another one.
```
ruby --jit-wait --disable-gems -e '1000.times { break if /a/ =~ "ab" && !$~[0] }'
``` -
- A mini reproductive command is here.
```
ruby --jit-wait --disable=gems -e '1000.times { a, b = nil }'
``` - 03:20 AM Revision 4c777ac9 (git): Revert "get rid of a compiler warning of VC"
- Partially revert r63820. mjit.c seems to have different idea of
"pid" type/size than the rest of Ruby on win32.
As noted in [ruby-core:87794], this seems to break Greg's build.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63856 b... - 03:02 AM Revision 44fc3d08 (git): unrevert r63852 but keep SIGCHLD path disabled for win32
- Reading win32/win32.c waitpid implementation, maybe waitpid(-1, ...)
on that platform will never conflict with mjit use of waitpid.
In any case, I've added WAITPID_USE_SIGCHLD macro to vm_core.h
so it can be easy for Linux/BSD users to ...
07/04/2018
-
11:20 PM Bug #14884: msys2 mingw32 'rake' has unexpected text at top of file
- > Try changing line 4 from `};{#` to `};{ #` (add a space before #). Worked for me in a Git bash shell running `rake --help`.
> ...
I'm using msys2's bash (with ruby installed via msys2's package manager). Putting a space there (or ind... -
- GardenTools (Garden Tools) wrote:
> Error message is
> ...
Try changing line 4 from `};{#` to `};{ #` (add a space before #). Worked for me in a Git bash shell running `rake --help`.
I don't know what shells you and @nobu use... -
- Error message is
/c/Ruby25/bin/rake: line 4: {#: command not found
-
- As it was reverted on r63852 by @naruse, assigning this back to Eric. I'm guessing it was reverted since it was not portable mainly for Windows. Whenever you commit things, you can check at least http://mswinci.japaneast.cloudapp.azure.c...
-
- I have nothing useful to add to the issue at hand as I only very
rarely use Net:: directly (my use cases usually start and end with
open-uri which I think is great) - the reason I comment here is to
suggest to you to add your feature ... -
08:43 PM Feature #14896 (Open): Expose Net::HTTPResponse socket for integration with libraries expecting an IO
- As several stdlib libraries support reading from IO-compatible objects (e.g. Zlib::GzipReader, Gem::Package::TarReader), would you be open to making the socket of a Net::HTTPResponse public?
This would enable such use cases as streami... -
- Constants defined in the namespace of a class or a module are named on the fly:
```ruby
class X; end
X.const_set(:Y, Module.new) # => X::Y
module M; end
M.const_set(:N, Module.new) # => M::N
```
Constants defined on a singleto... -
- @ioquatix nothing prevents from adding new method:
`Thread.create(scheduler: :current, args: [block_arg1, block_arg2]){|arg1, arg2| ... }`
Just remark: make test example to use `Fiber.transfer` for the reason above. -
- @funny_falcon I don't think it's easy to change `Thread.new(...)` because arguments are passed to the thread block.
However, if you want an API which doesn't introduce any new class, please check out https://bugs.ruby-lang.org/issues/... -
- Considering implementation, `Fiber.transfer` should be used, not `Fiber.resume`+`Fiber.yield`,
ie exclusively use `fiber_switch(..., ..., ..., 0);` and never use `fiber_yield(..., ...)`+`fuber_resume(...,...,...)`.
Reason: to allow nes... -
- Just remark: make test example to use `Fiber.transfer`.
If patch will be accepted, example will be copied, and usage of `Fiber.yield` is not composable with Enumerator. -
- Some of the benefits of this PR are:
- Makes it possible to implement different models for concurrency.
- Easy to implement by other Ruby implementations.
- Baseline interface for implementing scheduling, on which other schedulers c... -
- @normalperson that's awesome, great effort! I really appreciate you taking these PRs seriously and the effort you are putting into it.
I'd love to have a systematic comparison. Honestly, the port scanner isn't a great benchmark becaus... -
- * signal.c (signals): pack signal names instead of references.
* signal.c (signm2signo): also reject too long signal name.
[ruby-core:87767] [Bug #14893]
reapply r63841 and r63842, which are unrelated to r63758 but had been
wrongly r... -
- @naruse
I'm sorry, maybe I should have been clearer. This isn't the same as r63839.
For platforms that don't support ENV['TZ'], the code in the test file doesn't yield to the blocks passed by all the tests, as you noticed with the... -
- Fixed by r 63839
- 03:08 PM Revision 36f3e5ec (git): * 2018-07-05
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63853 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- The change is unstable on Windows. Please re-commit it when it correctly
supports Windows.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63852 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- `Gem::Package::TarWriter#add_file_signed` expects to fallback to
`Digest::SHA512`, and `digest.respond_to? :name` or not.
So lib/rubygems/security.rb should use same logic for
`Gem::Security::DIGEST_ALGORITHM` and `Gem::Security::DIGEST_... -
- Some tests are already skipped, but some tests are not.
So latter should be skipped too.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63850 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- This reverts commit r63848.
It breaks tests.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63849 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Time.utc uses timegmw() and it uses leap second information.
If the given time is larger than known_leap_seconds_limit, it calls
find_time_t, which uses localtime(3) and calls stat(2) in it.
This patch avoid it by setting known_leap_sec... -
04:39 AM Bug #14863: Array#join with empty array returns empty string always in US-ASCII encoding
- nobu (Nobuyoshi Nakada) wrote:
> May I close this?
Yes, please do.
I agree that US-ASCII is the greatest common denominator for source encodings, and there isn't any program that would fail (except if somebody checks for UTF-8 or ... - 01:26 AM Revision 4b20479f (git): internal.h (rb_warning_string): use PRINTF_ARGS
- Let compilers do a little extra validation for us
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63847 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 01:26 AM Revision d9dac20b (git): thread_pthread.c: pass rb_vm_t to timer_thread_sleep
- I love `container_of' for generic data structures, but
in this case it's unnecessary and slightly harder-to-read.
This will make "Timeout in VM" slightly easier-to-read:
https://bugs.ruby-lang.org/issues/14859
git-svn-id: svn+ssh://ci....
07/03/2018
-
- * vm.c (core_hash_merge_kwd): simplified to merge the second hash
into the first hash.
* compile.c (compile_array): call core#hash_merge_kwd with 2
hashes always, by passing an new empty hash to at the first
iteration.
git-svn-id... -
11:53 PM Bug #14894 (Closed): Segfault loading iseqs
- Follow up on https://github.com/Shopify/bootsnap/issues/172
In particular:
https://gist.githubusercontent.com/SamSaffron/1fe47b86374823fb620a6a29c83538ec/raw/c7e7b8f9bd605ab066bd66258ec76eab6df7fdff/dump.txt
Seeing a segfault... -
- I support adding `String#byteslice!`. I've been using `String#byteslice` in custom IO-like objects that implement `IO#read` semantics, as the strings I work with don't necessarily have to be in binary encoding (otherwise I'd just use `St...
- 09:29 PM Revision 8d004ee5 (git): mjit.h: avoid signed pointer comparisons (fix for 32-bit)
- ptrdiff_t is a signed type, use uintptr_t instead for unsigned
comparisons.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63844 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 03:25 PM Revision 1b812422 (git): * 2018-07-04
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63843 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- signal.c:69:10: error: variably modified ‘signm’ at file scope
char signm[LONGEST_SIGNAME + 1];
^
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63842 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset trunk|r63841.
----------
signal.c: packed signals
* signal.c (signals): pack signal names instead of references.
* signal.c (signm2signo): also reject too long signal name.
[ruby-core:87767] [Bug #14893] -
12:28 PM Bug #14893 (Closed): Global buffer overflow in signm2signo of signal.c.
- Found some memory error with address sanitizer:
```
==29152==ERROR: AddressSanitizer: global-buffer-overflow on address 0x55fb96d91983 at pc 0x7f80615106c6 bp 0x7fff6ee86480 sp 0x7fff6ee85c28
#1 0x55fb96aee1e7 in signm983 thread... -
- * signal.c (signals): pack signal names instead of references.
* signal.c (signm2signo): also reject too long signal name.
[ruby-core:87767] [Bug #14893]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63841 b2dd03c8-39d4-4d8f-98ff... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63840 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- May I close this?
-
- * [Feature #13050] Readline: expose `rl_completion_quote_character` variable
* [Feature #14850] Add official API for setting timezone on `Time`
* [Feature #14869] Proposal to add `Hash#===`
* [Feature #14877] Calculate age in `Date` c... -
- It's polyglot code for sh and cmd.exe.
What error do you have with it actually? -
- I recognize that. I'm currently taking time for RubyBench project but planning to take a look after that. Please ignore the -DMJIT_FORCE_ENABLE=1 until I fix it.
-
- r63763 ("give up insn attr handles_frame") breaks -DMJIT_FORCE_ENABLE=1
(commit 6b534134a78e3e43c344682c3585e1abab015216 in git)
Related [Bug #14809] [Bug #14834]
Is this being worked on? I mentioned it in [ruby-core:87674],
"mak... -
11:22 AM Feature #14869: Proposal to add Hash#===
- > 空の場合は Enumerable#all? と類似の話なので、 true のほうがいいんじゃないでしょうか。
なるほど、参考になります。
> ...
あーこれは確かに問題になりそうですね。対策を考えてみたいと思います。
ありがとうございます。 -
- For example Solaris
https://rubyci.org/logs/rubyci.s3.amazonaws.com/unstable10x/ruby-trunk/log/20180703T091803Z.fail.html.gz
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63839 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- ref r63801, r61199
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63838 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 09:06 AM Revision 18c79781 (git): vm.c (rb_source_location): check for NULL cfp->iseq
- rb_vm_get_sourceline returns zero if cfp->iseq is NULL,
so rb_iseq_path should not try to follow NULL cfp->iseq,
either.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63837 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 08:30 AM Revision 45cdc834 (git): thread_pthread.c (native_thread_destroy): clear native TSD pointer
- mwrap <https://80x24.org/mwrap/> interposes malloc functions and
checks for GVL existence to determine Ruby source locations of
malloc calls. pthread_getattr_np (from get_stack) may call
realloc to get the CPU set size; so when using th... -
More examples:
~~~ ruby
Pathname.new('/a').join('c', 'b').to_s #=> "/a/c/b"
Pathname.new('/a').join('/c', 'b').to_s #=> "/c/b"
Pathname.new('/a').join('/c', '/b').to_s #=> "/b"
~~~
Why it does this?
because `"/c"` means star...-
- This behaviour surprised me too. Is it expected that the information from "/a"
is lost? I have no idea but the documentation does not mention this; perhaps
the above could be added as an example OR the behaviour changed (or both).
I... -
02:37 AM Bug #14891 (Closed): Pathname#join has different behaviour to File.join
- ~~~ ruby
Pathname.new('/a').join('/b').to_s
# => "/b"
File.join(Pathname.new('/a'), '/b').to_s
# => "/a/b"
~~~
in my case `'/b'` was in a variable and it wasn't immediately obvious why it wasn't working when I moved to use Pathna... -
- It may raise an error in a certain security configuration.
It is very likely to trigger a segmentation fault if `getcontext()` failed silently
and we just let it keep going.
Related to https://bugs.ruby-lang.org/issues/14883
[Fix GH-190... -
- * template/fake.rb.in: removed duplicated options in RUBYOPT.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63834 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * spec/ruby/command_line/rubyopt_spec.rb: skip -v in RUBYOPT
examples when CROSS_COMPILING is set by fake.rb. the version
number by -v is printed before loading libraries by -r options,
so setting RUBY_DESCRIPTION in fake.rb has n... -
- cflags, cppflags, and cxxflags are additional flags to auto
configured flags, and ignored when the their respective autoconf
default variables are given.
[ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63832 b2dd03c8-39d4-4d... - 12:29 AM Revision 54707221 (git): test/ruby/test_signal.rb: speedup some test by disabling RubyGems
- test_signame_delivered is reduced from 1.0s to 0.1s
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63831 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
07/02/2018
-
- nobu@ruby-lang.org wrote:
> Since r63758, `rb_waitpid()` hangs up frequently (not always) on macOS (darwin17).
> Seems `SIGCHLD` is not delivered.
I guess sigsetjmp in r63803 did not help? I don't think it
would, once a thread st... -
- k0kubun / usa / nobu: can you check the return value handling of
_WIN32 waitpid_sys/WaitForSingleObject in process.c as of r63797?
It needs to retry on the timeout==0 (WNOHANG) case.
Thanks.
- 11:14 PM Revision a7b874b3 (git): * 2018-07-03
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63830 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 11:14 PM Revision ca37c5d4 (git): process.c (waitpid_nogvl): start timer thread polling for lossy SIGCHLD
- For systems with lossy SIGCHLD, an infinitely sleeping timer
thread needs to be aware of rb_waitpid callers in the first
place before it can check and reset polling status.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63829 b2dd03c... -
- Hi Greg,
> Assuming you're running from a cmd/ps shell. Looking at the first few lines of the script, it may also run with a bash shell...
The difficulty is running from bash, which finds 'rake' (and becomes confused when it tries... -
- You probably meant "... call #delete_at in the background".
I think there could be a performance issue for big arrays if calling `delete_at`. For example `delete_if{true}` would be moving `O(n^2)` elements (for no reason!) while `dele... -
- I didn't think to enumerators...
Yes, #delete_at is a better choice than #delete, didn't think to it neither...
But #delete and #delete_if methods seem to not call #delete_at neither. Please let me reword the proposal: may #delete ... -
- Do you mean delete_if and reject! should call *delete_at*?
Calling #delete doesn't make sense, as @Hanmac showed above.
In general, I would not recommend to inherit from Array, there are many pitfalls.
It is much simpler and bette... -
- I don't think there is a way:
~~~ ruby
a = [1,2,3,2,5]
a.delete_if {|x| x == 2}
a #=> [1,3,5]
~~~
this might call delete, but what about this?
~~~ ruby
a = [1,2,3,2,5]
a.delete_if.with_index { |x,i| x == 2 && i == 1 }
a #=... -
- Hmm, let me make it a feature request so that we can consult matz.
-
07:21 PM Bug #14883: Ruby 2.5 Fails to Build on PowerPC 32-bit (BE)
- Bingo. It is also an issue for the systemd-nspawn container indeed.
systemd-nspawn have used a libseccomp whitelist instead of a blacklist by default since last year.
`rt_sigprocmask()` & `sigprocmask()` are on the whitelist, but `sw... -
- contd.
When it comes to a glibc function `getcontext()`, 'the root issue' of the bug, we can check it in this way:
~~~c
// cat lion_getcontext.c
#include <ucontext.h>
#include <errno.h>
int main() {
ucontext_t c;
if (... -
- I did some further investigation. It is a little bit longer story than I thought.
Firstly, the callstacks in Ruby shows that:
~~~
...
/root/ruby/lib/rdoc/parser/ripper_state_lex.rb:594:in `parse'
/root/ruby/lib/rdoc/parser/ripper_... -
- naruse (Yui NARUSE) wrote:
> Your build/SEGV log doesn't have the essential part of the crash log.
Bai's friend here. The missing part of the crash log: https://pastebin.aosc.io/paste/fBNM~Lr~C4vfsvHCpmPJOA
-
- Your build/SEGV log doesn't have the essential part of crash log.
-
06:15 AM Bug #14883: Ruby 2.5 Fails to Build on PowerPC 32-bit (BE)
- mingcongbai (Mingcong Bai) wrote:
> When building Ruby 2.5.1 on a PowerPC 32-bit (Big Endian) host, the build fails as it generates rdoc - segmentation fault. The build log is as follows:
> ...
This issue is introduced with the followi... -
- Should the test file test/ruby/test_time_tz.rb be skipped on Windows?
ruby-loco has some additional reporting features, and until r63823, none of the asserts were actually running (see with_tz definition).
Since no asserts are run,... -
- You can already use `TracePoint` to "catch" the calls to eval:
```
TracePoint.trace(:c_call) do |tp|
if tp.callee_id == :eval
p "Eval called!"
end
end
```
This won't give you access to the arguments though. I'm not su... -
02:32 PM Bug #14882 (Third Party's Issue): OpenSSL::PKey::*.new does not resume after spurious signals
- Thank you for the report. I made a GitHub PR that fixes the bug at the GitHub repo (add .patch to the URL for "git format-patch"-style diff):
https://github.com/ruby/openssl/pull/205 -
01:51 PM Bug #14876 (Closed): /bin/sh: bad substitution since r63679 on Solaris
- Applied in changeset trunk|r63828.
----------
Makefile.in: ARCH_FLAG may contain "-m64" etc.
* Makefile.in (mjit_config.h): ARCH_FLAG may contain "-m64", "-m32",
"-march=i486" and so on, but the change made with r63232 only
support... -
- * Makefile.in (mjit_config.h): ARCH_FLAG may contain "-m64", "-m32",
"-march=i486" and so on, but the change made with r63232 only
supports "-arch AAA -arch BBB" mainly used on Mac OS X.
To solve the issue, ARCH_FLAG is parsed and ... -
- ko1@atdot.net wrote:
> https://bugs.ruby-lang.org/issues/14857
> Collected information:
>
> * gen (GC count at allocation)
> * file (C file name)
> * line (C line number at the file)
Btw, I posted a complementary idea, "mwrap"... -
- Thanks!
-
12:31 AM Bug #14846: Backport fixes for Kernel#warn(uplevel:)
- ruby_2_5 r63819 merged revision(s) 61525,61527,63062,63650.
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63827 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
-
- ruby_2_5 r63826 merged revision(s) 63304.
-
- io.c: workaround for EPROTOTYPE
* io.c (internal_write_func, internal_writev_func): retry at
unexpected EPROTOTYPE on macOS, to get rid of a kernel bug.
[ruby-core:86690] [Bug #14713]
* ext/socket/in... -
- mifrill (Aleksey Strizhak) wrote:
> but the problem is, the behavior is different for all three objects.
`Date` doesn't have time zone, and `Date#to_time` borrows the system timezone.
> ...
`Date` is a discrete library, but `Time`... -
- ruby_2_5 r63825 merged revision(s) 63696.
-
- variable.c: fix receiver on private constant
* variable.c (rb_const_search): fix NameError :receiver attribute
on private constant, should raise with the included module, not
the ICLASS.
git-svn-id: svn+ssh:... -
- 空の場合は `Enumerable#all?` と類似の話なので、 `true` のほうがいいんじゃないでしょうか。
また、再帰的なハッシュに対して使用したときに無限再帰にならないようにする必要がありそうです。 -
- Applied in changeset ruby_2_5|r63824.
----------
merge revision(s) 62731,62735: [Backport #14495]
Bug Fix Enumerator::Lazy#uniq state for multiple call
* enumerator.c (lazy_uniq_i): create new hash for each calls.
[Fix GH-1820]
... -
- ruby_2_5 r63824 merged revision(s) 62731,62735.
-
- Bug Fix Enumerator::Lazy#uniq state for multiple call
* enumerator.c (lazy_uniq_i): create new hash for each calls.
[Fix GH-1820]
Currently
2.5.0-preview1 :001 > arr = (0..100).lazy.uniq{|i| i... -
- Applied in changeset trunk|r63823.
----------
time.c: [DOC] Time#localtime
* time.c: state that Time#localtime does nothing when nothing
changes. [ruby-core:87675] [Bug #14880] -
- * time.c: state that Time#localtime does nothing when nothing
changes. [ruby-core:87675] [Bug #14880]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63823 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * ruby.c (cmdline_options_init): set up mjit.on flag by
MJIT_FORCE_ENABLE in the initialization function.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63822 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- - [Bug #14887] Array#delete_if does not use #delete (shyouhei)
- Is it by design, or a bug? - 01:50 AM Revision e2b7363e (git): * 2018-07-02
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63821 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 01:50 AM Revision dea84229 (git): get rid of a compiler warning of VC
- * mjit.c (exec_prcess): use PRI_PIDT_PREFIX for pid.
* win32/Makefile.sub (PRI_PIDT_PREFIX): force to "I".
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63820 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- error.c: negative uplevel
* error.c (rb_warn_m): negative uplevel is not allowed.
[ruby-core:84568] [Bug #14262]
error.c: limit depth
* error.c (rb_warn_m): limit backtrace depth to reduce objects to
... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_5@63818 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- ruby_2_5 r63817 merged revision(s) 63054,63055,63056.
-
- test_time_tz.rb: Kiritimati tzdata fix
* test/ruby/test_time_tz.rb (TestTimeTZ#test_pacific_kiritimati):
fix the expected data at the skip of New Year's Eve 1994.
[Bug #14655]
test_time_tz.rb: Kiriti... -
- It would be nice to expose the class and method/function name if possible. In the case of code coverage, it would allow user to explicitly ignore functions, because we can traverse the AST and ignore every line of the function.
- NODE...
07/01/2018
-
- I don't have anything overly helpful to the discussion to add; but I wanted
to add one thing to this:
> We usually write an optional argument in a line: def print_summary(out = $stdout)
While I concur in general, I myself have exp... -
- I'm unsure about your assumption. If you can insert a newline freely, you may want to write the following code:
```
30 1| def print_summary(
31 | output = (
32 0| $stdout
33 | ... -
- If this is by design, please feel free to close, but as it stands there is no way to check if optional argument evaluation occurred or not. For code coverage, this is a limitation.
-
- I have the following code.
```
30 1| def print_summary(
31 0| output = $stdout
32 | )
```
In the margin is the line number followed by the number of hits according to `:line` trace point.
I feel like line 31 sh... -
08:12 AM Revision b3fb892d (git): NEWS: add NEWS entry about Kernel#then [Feature #14594]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63816 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- lib/csv.rb was divided into lib/csv/*.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63815 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63814 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63813 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 01:38 AM Revision fe398f87 (git): * 2018-07-01
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63812 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- [Bug #14246]
[ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63811 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I'm working on an improved coverage library for Ruby which can compute coverage of things like ERB templates where the source file is not Ruby code but is translated to Ruby code using `eval`.
When using the line trace point, the path...