Activity
From 07/21/2021 to 07/27/2021
07/27/2021
-
09:48 PM Bug #10902: require("enumerator") scans LOAD_PATH 2x on every invocation
- If we want to allow loading an `enumerator.rb` library, then we cannot avoid the scan of the load path for *.rb files. However, we can avoid scanning the load path twice, once for `enumerator.rb` and once for `enumerator.so`, since we a...
-
09:11 PM Feature #18042: YARV code optimization
- I think you'd have to do a lot of work, to even get the basic benefits of this approach. So they may not want what is likely to be a very large volume of code for what is likely to be a small benefit.
-
07:29 PM Feature #18042: YARV code optimization
- But way MRI team don't want to borrow the developments of other ruby implementations? because it's hard to support?
-
- Yes these are all optimisations that compilers can do and you can see them being done in implementations of Ruby like TruffleRuby, but it requires a complex system of 'deoptimisation' to be able to go from the optimised code back to the ...
-
- @nobu Yes you are right, but I want to say that we don't need temp object in this case, so wy can't we do some optimizations? or one more case: code like this `[1,2,3,4,5,6].select(&:even?).count` will allocate new array only for return ...
-
11:15 AM Feature #18042: YARV code optimization
- Hanmac (Hans Mackowiak) wrote in #note-6:
> motoroller (Iskandar Gohar) wrote:
> ...
And the second should be `_1.method1.method2`. -
07:11 AM Feature #18042: YARV code optimization
- motoroller (Iskandar Gohar) wrote:
> ```ruby
> ...
there are differences with the code:
* first lets say that array is `[a, b]`
* the first example would call x = a.method1, y = b.method1 and then `x.method2, y.method2`
* the seco... -
06:12 AM Feature #18042 (Feedback): YARV code optimization
- > as I see mjit only translates yarv into C
It's interesting that many people seem to think it's what MJIT does. We indeed copy-paste the original C code for most instructions, but it's not for optimizations but just for ease of mainten... -
07:36 PM Bug #14500 (Closed): Missing Regexp documentation and clarification on behavior of \K for edge case
- Applied in changeset commit:git|4fc9ddd7b6af54abf88d702c2e11e97ca7750ce3.
----------
Update Capturing and Anchors sections of regexp documention
Document that only first 9 numbered capture groups can use the \n
backreference syntax. D... -
- Sundeep (Sundeep Agarwal) wrote:
> **Anchors** section (https://ruby-doc.org/core-2.5.0/Regexp.html#class-Regexp-label-Anchors)
> ...
The `\K` behavior at the end of a regexp is a bug that has already been filed upstream in Onigmo: htt... -
- Document that only first 9 numbered capture groups can use the \n
backreference syntax. Document \0 backreference. Document \K anchor.
Fixes [Bug #14500] -
- Applied in changeset commit:git|35e467080ca35a9a129e95f802f102c3bc0a81b3.
----------
Make Float#floor with ndigits argument handle error
The previous implementation could result in a returned
float that is 1/(10**ndigits) too low. Fir... -
05:48 PM Bug #18018: Float#floor / truncate sometimes result that is too small.
- jeremyevans0 (Jeremy Evans) wrote in #note-5:
> > A correct algorithm seem to be to rely on `Rational#floor`:
> ...
My bad 🤦♂️ Rational#floor makes sense for rationals, but is not what we can use. Not sure how I got confused.
> I t... -
- The previous implementation could result in a returned
float that is 1/(10**ndigits) too low. First try adding
one before dividing, and if that results in a value that is
greater than the initial number, then try the original
calculatio... -
- Applied in changeset commit:git|a6dc48f86a8ee7b1f713105b5c7136824d6b5720.
----------
Fix infinite loop in ensure after NoMemoryError
VM patch from wanabe.
Test based on example from buzztaiki (Taiki Sugawara).
Test fails when compile... - 05:48 PM Revision 338b604b (git): * 2021-07-28 [ci skip]
-
- VM patch from wanabe.
Test based on example from buzztaiki (Taiki Sugawara).
Test fails when compiles with -DRUBY_DEBUG, as that can
can use rb_bug instead of NoMemoryError, which doesn't
allow testing this case. Test also fails on Min... -
- I submitted a pull request upstream to fix this: https://github.com/ruby/date/pull/35. However, I'm not sure if the consistency benefit is worth the backwards incompatibility cost.
-
05:08 PM Feature #18047: TracePoint: Add event type for constant access
- Is it only on constant reads or also on constant writes/assignments?
Access is a pretty generic term and I would think it means both, but then is there a way to differentiate from the TracePoint event?
Adding docs to the TracePoint class... -
01:46 AM Feature #18047 (Feedback): TracePoint: Add event type for constant access
- Hi there
We've made a PR to add new :constant_access TracePoint event. https://github.com/ruby/ruby/pull/4683
Burke Libbey made some first steps on this idea a few years ago #13133 and we improved his patch based on the comment by ko1.... -
03:52 PM Bug #18049: Cannot build ruby-3.0.2 on NetBSD
- First, it should be clarified that it is the already installed `/usr/local/bin/ruby` (version 3.0.1p64) that segfaults here.
I guess that the source comes directly from a Git repository, as `prelude.rbinc` should be prebuilt in a rele... -
01:49 PM Bug #18049 (Feedback): Cannot build ruby-3.0.2 on NetBSD
- I get random segfaults when attempting to build on NetBSD
```sh
uname -a
NetBSD juggler.panix.com 9.1 NetBSD 9.1 (PANIX-STD) #2: Thu Mar 18 17:46:23 EDT 2021 [email protected]:/misc/obj64/misc/devel/netbsd/9.1/src/sys/arch/amd... -
- ```
NetBSD juggler.panix.com 9.1 NetBSD 9.1 (PANIX-STD) #2: Thu Mar 18 17:46:23 EDT 2021 [email protected]:/misc/obj64/misc/devel/netbsd/9.1/src/sys/arch/amd64/compile/PANIX-STD amd64
Configuration summary for ruby version 2.7.... -
- This error can occur on Windows for certain filenames on certain
code pages.
Fixes [Bug #14591]
https://github.com/ruby/find/commit/0a474d1027 -
11:56 AM Feature #12675 (Closed): Code simplification and readability for prime_division method in prime.rb
- prime.rb was extracted from ruby repository to https://github.com/ruby/prime. If you still interested in this, Can you file it to ruby/prime repo? Thanks.
-
- You can find it like `lib/ruby/3.1.0/x86_64-darwin20/json/ext/parser.bundle`.
-
-
- Applied at https://github.com/ruby/cgi/commit/724c9e0e0acf29f39471393b1b67239760d8ce14
-
09:20 AM Bug #18028 (Closed): test/ruby/enc/test_emoji_breaks.rb does not deal with Unicode ranges in file emoji-sequences.txt
- Completed with commit 26b1e6fca8.
-
- Detect Unicode ranges and loop over them.
This fixes issue #18028. -
-
- Completed with revisions 01e98d8785 and 3010b40477.
-
-
-
- Deal with the issue that the emoji files in emoji/13.1 have Unicode
Emoji version 13.1, but at the same time the files in 13.0.0/ucd/emoji
are still at Emoji version 13.0. Specifically:
- Add a version attribute to TestEmojiBreaks::Break... -
-
- https://github.com/ruby/uri/commit/bbf8b44dba
-
- * Using a module to map scheme name to scheme class, which also works with Ractor.
* No constant redefinition, no ObjectSpace, still fast lookup for initial schemes.
https://github.com/ruby/uri/commit/883567fd81 -
- * This reverts commit 1faa4fdc161d7aeebdb5de0c407b923beaecf898.
* It has too many problems, see https://github.com/ruby/uri/pull/22 for discussion.
https://github.com/ruby/uri/commit/b959da2dc9 -
07:54 AM Revision a288c21a (git): [ruby/uri] Fix to support Ruby 3.0 Ractor
- https://github.com/ruby/uri/commit/1faa4fdc16
-
07:25 AM Bug #18048: Thread#join can break with fiber scheduler unblock fails or blocks.
- The PR is 050a89543952a2c9e7c9bc938f4fdb538f6c9278 followed by 13f8521c630a15c87398dee0763e95f59c032a94
-
07:22 AM Bug #18048 (Closed): Thread#join can break with fiber scheduler unblock fails or blocks.
-
- Thank you for creating the pack for backport. I see the PR was basically backporting 050a89543952a2c9e7c9bc938f4fdb538f6c9278 partially. I will try to merge it.
-
- In addition to https://bugs.ruby-lang.org/issues/17666 we found several more cases that need to be addressed.
Fix potential hang when joining threads.
If the thread termination invokes user code after `th->status` becomes
`THREAD_... -
-
-
-
-
- 06:23 AM Revision 13f8521c (git): Fix potential hang when joining threads.
- If the thread termination invokes user code after `th->status` becomes
`THREAD_KILLED`, and the user unblock function causes that `th->status` to
become something else (e.g. `THREAD_RUNNING`), threads waiting in
`thread_join_sleep` will ... -
-
04:41 AM Revision 2de7c29b (git): test/ruby/test_parse.rb: Prevent an "unused variable" warning
-
-
- Consume the VM stack more, to make the target object get GCed with
more probability during suppressing the warning. -
02:37 AM Bug #17429: Prohibit include/prepend in refinement modules
- I've changed the behavior and have added tests in https://github.com/shugo/ruby/pull/3/commits/3aaca9217f958640495c14ac11b08948960d1f30
- 12:57 AM Revision 5b27c1a1 (git): * 2021-07-27 [ci skip]
-
12:25 AM Revision bd8daa85 (git): [rubygems/rubygems] Remove the warning for not defining a gem server source
- Bundler has deprecated gemfiles without a global source and this feature
is now obsolete. `Bundler::Definition#has_rubygems_remotes?` is removed
because it's not used anymore.
https://github.com/rubygems/rubygems/commit/d29dd2cb7b -
- `check_rubygems_source_safety` is responsible for:
1. if there are multiple global sources
- for bundle 3.x raise an error
- for bundle 2.x print a warning
2. print a warning if there is no explicit global source
The second respons... -
- Raise a warning when parsing a Gemfile and it doesn't have a global source. Gemfiles like this, specially now that rubygems sources are are no longer merged into a single source for security, are very confusing because they generate a di...
-
- This is in preparation for deprecating source-less gemfiles.
https://github.com/rubygems/rubygems/commit/d6493fa3e2 -
- This method is created to tell whether any global source exist in the object or not and it will be used by `Bundler:Dsl` to print a warning if no global source has been defined in the Gemfile.
https://github.com/rubygems/rubygems/commit... -
- This method is created to tell whether any remote exist in the object or not and it will be used by `Bundler:SourceList` to tell if a global source has been defined implicitly or not.
https://github.com/rubygems/rubygems/commit/47e3ff0e47 -
12:25 AM Revision 095b5bbc (git): [rubygems/rubygems] Lazily load shellwords
- https://github.com/rubygems/rubygems/commit/ef6388656a
-
- https://github.com/rubygems/rubygems/commit/fe723c0f7f
-
- https://github.com/rubygems/rubygems/commit/8bf41f53cc
-
- https://github.com/rubygems/rubygems/commit/e5532ef886
-
- The is the previous intentional behaviour until
https://github.com/rubygems/bundler/commit/ca0676cb1c638e0b9747ea8c18f28adf82cc01de.
In my opinion, that previous behaviour was better and should be
restored, because we want our users to ... -
- We'll be removing the warning about no gem sources, so this spec will no
longer test that warnings are hidden by `--quiet`.
Test that in another way so that we don't lose the coverage when we
drop the specific warning about no gem serve... -
- Mitigate the security risk:
https://devcraft.io/2021/01/07/universal-deserialisation-gadget-for-ruby-2-x-3-x.html
https://github.com/rubygems/rubygems/commit/141c2f4388 -
- After reading [this blog
post](https://blog.rubygems.org/2011/08/31/shaving-the-yaml-yak.html),
published almost 10 years ago already, my understanding is that this
problem could come up in two ways:
* Rubygems.org serving corrupted gem... -
- A fresh `gem install` might not reproduce the exact `bundle install`
environment that originally caused the error. It also makes it harder
for the user to troubleshoot the error since she needs to run a separate
command.
Instead, show t... -
- For example, due to extension compilation issues.
https://github.com/rubygems/rubygems/commit/adbe55bb6a -
- To mimic built-in rubygems behaviour, only thing that should be
approximated is the lockfile version. Other alternatives like
`BUNDLER_VERSION` should be respected exactly.
https://github.com/rubygems/rubygems/commit/dbd667d4bc -
- Ruby 1.9.2 removed "." from LOAD_PATH for robustness and security reasons.
This code was introduced by https://github.com/rubygems/rubygems/commit/56fc830e19a573a5905eba7f4714ad1f21ed1927 commit
to helping users understand the issue and ... -
12:25 AM Revision 705b1bde (git): [rubygems/rubygems] Fix interrupt handling in Bundler workers
- The existing interrupt handling using `SharedHelpers.trap` fails when the previous
handler for a signal is not callable (for example, when it is the string "DEFAULT").
Instead, we now handle interrupts by aborting the process when worke...
07/26/2021
-
- Making `ruby/internal/intern/thread.h` depend directly on `ruby/io.h` doesn't work due to circular dependency issues. I have a commit that adds a separate include file that both `ruby/internal/intern/thread.h` and `ruby/io.h` include to ...
-
07:00 PM Bug #18046 (Closed): Missing include leads to build breakage of glib2 on head
- I was testing out my code on a CI and I noticed that the build broke with some compile errors. I reported it to the ruby-gnome project, and they think it's a missing include. Here's that bug: https://github.com/ruby-gnome/ruby-gnome/issu...
-
- This issue still exists in the master branch. I've submitted a pull request that uses a modified version of @mame's patch to fix this issue: https://github.com/ruby/ruby/pull/4682
-
- marcandre (Marc-Andre Lafortune) wrote in #note-4:
> > marcandre (Marc-Andre Lafortune) wrote:
> ...
I did miss that :). I agree with you that it is reasonable definition.
> A correct algorithm seem to be to rely on `Rational#floor`... -
05:44 PM Feature #18045 (Closed): Variable Width Allocation Phase II
# GitHub PR: https://github.com/ruby/ruby/pull/4680
# Feature description
Since merging the initial implementation in #17570, we've been working on improving the performance of the allocator (which was one of the major drawbacks ...-
- as I see mjit only translates yarv into C
-
- This is how the various JITs for Ruby work. You usually think of a JIT emitting machine code, but they could also emit optimised YARV instructions.
-
-
06:32 AM Bug #18044: unusual behavior with Arabic string in a Hash
- Thanks ! Indeed removing that invisible character fixes the issue. I wonder if this error could be made more obvious?
-
06:24 AM Bug #18044: unusual behavior with Arabic string in a Hash
- If I'll copy-paste your example literally, and then...
```ruby
foo.keys.first.chars
# => ["a", "r", "a", "b", "i", "c", ""]
foo.keys.first.chars.last.unpack('U') # => 8206
8206.to_s(16) # => 200e
```
It is [U200e](https://www.com... -
- Hey all,
I found a potential bug where looking up a key that contains an arabic string always returns "nil" instead of the string. To reproduce, run the following script
```ruby
foo = {"arabic" => "ٱلتَّوْبَة"}
p foo.keys # => [... -
-
- In other words, Refinement#import copies methods but does not attempt to change anything lexical, except which refinements are applied to these methods.
-
- Yes, I think that's completely fine.
In code, this should work:
```ruby
module Shared
A = 1
def foo
A
end
end
refine SomeClass do
import Shared
end
SomeClass.new.foo # => 1
```
And this should not:
```r... -
- Eregon (Benoit Daloze) wrote in #note-23:
> shugo (Shugo Maeda) wrote in #note-22:
> ...
For me, it's surprising if the imported methods cannot access constants of the refinement.
However, constant assignments in refine block define cons... -
- And revert 24e5f1c982966c379220b1bbb26b4e0320180fa1, pepending to
Kernel did not affect the top level methods before 3.0. -
- `DebugSystem#system` is prepended in vcs.rb and defaulted to
`exception: true`.
07/25/2021
-
- Applied in changeset commit:git|070557afc4ca83876b951fe090806b59e3867ae5.
----------
Distinguish signal and timeout [Bug #16608] - 08:09 PM Revision 165a098f (git): * 2021-07-26 [ci skip]
-
-
-
- The same as #17988 (and others related to Ethon on M1 Mac). Should be closed as a third party's issue.
-
10:25 AM Bug #18043 (Third Party's Issue): pod install failed on iMac (24-inch, M1, 2021) macOS Big Sur Version: 11.4 (20F71)
- % pod install
Analyzing dependencies
/Library/Ruby/Gems/2.6.0/gems/ffi-1.15.3/lib/ffi/library.rb:275: [BUG] Bus Error at 0x0000000104838000
ruby 2.6.3p62 (2019-04-16 revision 67580) [universal.arm64e-darwin20]
-- Control frame inform... -
- jeremyevans0 (Jeremy Evans) wrote in #note-3:
> This issue also goes the other direction:
> ...
Well, that `f` *is* the same float (291.4), I don't understand your point, sorry.
> marcandre (Marc-Andre Lafortune) wrote:
> ...
Maybe... -
- Thanks for working on this and coming up with a 100% backwards compatible fix! <3
-
- Applied in changeset commit:git|b360588cd3cbac5fb4f004aa53a8fdc715906719.
----------
Sort feature index arrays by the priority of file types [Bug #15856]
When looking for libraries to load with a feature name without
extension, `.rb` f... -
11:58 AM Revision 8897098b (git): Update bundled_gems
- 11:07 AM Revision 02facb86 (git): * 2021-07-25 [ci skip]
-
- Fixed leaked global symbols
---
gc.c | 14 +++++++++-----
vm_insnhelper.c | 7 ++++---
vm_insnhelper.h | 2 +-
vm_sync.c | 4 ++--
vm_sync.h | 6 ++++--
... -
- Reduced GitHub Actions
Moved leaked-globals before check.
---
.github/workflows/macos.yml | 4 +++-
.github/workflows/ubuntu.yml | 8 +++-----
2 files changed, 6 insertions(+), 6 deletions(-)
... -
- ruby_3_0 95dc88c88869541dd0eccafd14924d78c8d7f427 merged revision(s) 5f69a7f60467fa58c2f998daffab43e118bff36c.
-
- Co-authored-by: Samuel Williams <@ioquatix>
https://github.com/nagachika/ruby/pull/1/commits/2cee515f024f3295945f312cb6b052f972f9c93d -
-
07/24/2021
-
- When looking for libraries to load with a feature name without
extension, `.rb` files are given priority. However, since the
feature index arrays were not in that order of priority, but in
the order in which they were loaded, a lower pri... - 08:58 AM Revision 971d24cd (git): * 2021-07-24 [ci skip]
-
07/23/2021
-
- The algorithm for `Float#floor` with the `ndigits` argument is basically (omitting the overflow/underflow handling):
```ruby
class Float
def floor_ndigits(ndigits)
x = (10**ndigits).to_f
(self * x).floor / x
end
end
... -
-
- I've reviewed @nobu's patch and think that we should go with that approach. It avoids the backwards compatibility issues with @deivid's approach while still fixing the performance issue. I submitted a pull request for it: https://githu...
-
- I agree with @phluid61, plain path provided to `URI.parse` should not be treated as implicit `file` scheme. Such a change would break backwards compatibility for the methods mentioned, and I think `URI.parse('http://foo/bar').merge(URI....
-
- I don't think this is a bug. `Time` defaults to local time, and `DateTime` to UTC:
```ruby
Time.parse('2021-07-23')
# => 2021-07-23 00:00:00 -0700
DateTime.parse('2021-07-23')
#<DateTime: 2021-07-23T00:00:00+00:00 ((2459419j,0s... -
- Hi! Long period of time I think about programmatically code optimization for YARV. In compiled languages like C/C++ the compiler can do whatever it wants with the code and does for performance optimization. Firstly, ruby developers think...
-
- shugo (Shugo Maeda) wrote in #note-22:
> In the current implementation, the module in cref is replaced with the refinement like Module#dup, so constants in the imported module are not accessible from the copied methods.
Could you sho... -
-
-
- Applied in changeset commit:git|fc4dd45d0142221880d1b2c9b54dee0597be2b78.
----------
Show exception in finalizer [Feature #17798] -
-
-
-
- It can be optimized and handles Qnil properly.
-
07/22/2021
-
- applenotbanana (Apple Banana) wrote in #note-1:
> is it problem with excon?
Excon is a pure ruby gem, and the segfault appears to occur in openssl, so my guess would be no. In any case, Ruby 2.5 is no longer supported, Please try wi... -
08:36 PM Bug #18041: [BUG] Segmentation fault - ruby 2.5.8p224 (2020-03-31 revision 67882) [aarch64-linux]
- is it problem with excon?
-
- /usr/local/lib/ruby/2.5.0/openssl/buffering.rb:388: [BUG] Segmentation fault at 0x0020ffffaf56467c
ruby 2.5.8p224 (2020-03-31 revision 67882) [aarch64-linux]
-- Control frame information ----------------------------------------------... - 06:13 PM Revision fc78a089 (git): * 2021-07-23 [ci skip]
-
06:13 PM Revision 86e0eecc (git): [ruby/irb] Fix #256
- Support int that follow on symbeg in IRB
https://github.com/ruby/irb/commit/90cb27b1bd -
- https://github.com/ruby/ruby/pull/4670
-
- We already page the page of the zombie calculated. Don't recalculate the
page. -
- The call after it to `heap_page_add_freeobj` will set the flags.
-
-
06:41 AM Revision fc50b2ea (git): Remove unneeded function declarations
- 05:10 AM Revision e4b68ab7 (git): [DOC] Fixed the description of regexp alternations [ci skip]
-
-
-
- Thank you for your confirmation and adding a test!
I will create a patch for ruby_3_0 branch based on your pull request. -
- ruby_3_0 a215c6d0448764131cbbb48b476dc698b51c2273 merged revision(s) 119697f61e2b2b157816a8aa33aada5863959900,4a627dbdfd1165022fa9e716ba845e937b03773d.
-
- Thank you for reviewing. I will commit the patch soon.
- 02:48 AM Revision 186900d1 (git): * 2021-07-22 [ci skip]
-
- [Bug #18014] Fix rb_gc_force_recycle unmark before sweep
If we force recycle an object before the page is swept, we should clear
it in the mark bitmap. If we don't clear it in the bitmap, then during
sweeping we ... -
-
- len and *retlen are size_t since r22957
(commit:4de12b6ae9a7fc1e28ada4b62291a78f28ae7528).
07/21/2021
-
- `gc_verify_internal_consistency_` does not walk pages in the tomb heap
so numbers were off. This commit changes it to walk all allocated pages. -
05:07 PM Bug #18038 (Closed): Invalid interpolation in heredocs
- Applied in changeset commit:git|b940a453572b5c3ed5c0951647929e14f5843a7d.
----------
Fix interpolated heredoc
This fixes https://bugs.ruby-lang.org/issues/18038. The provided
reproduction showed that this happens in heredocs with doubl... - 05:06 PM Revision f5f70106 (git): * 2021-07-22 [ci skip]
-
- This fixes https://bugs.ruby-lang.org/issues/18038. The provided
reproduction showed that this happens in heredocs with double
interpolation. In this case `DSTR` was getting returned but needs to be
convered to a `EVSTR` which is what is... -
-
-
- Instead of scanning ruby_hexdigits.
- 08:40 AM Revision b2749e00 (git): * 2021-07-21 [ci skip]
-