Activity
From 12/10/2025 to 12/16/2025
Today
-
08:00 PM Revision 5e27581c (git): ZJIT: Use rb_zjit_writebarrier_check_immediate() instead of rb_gc_writebarrier() in gen_write_barrier()
- * To avoid calling rb_gc_writebarrier() with an immediate value in gen_write_barrier(),
and avoid the LIR jump issue. -
-
-
-
-
08:00 PM Revision 094418a6 (git): gc.h: Reintroduce immediate guard in `rb_obj_written`
- This guard was removed in https://github.com/ruby/ruby/pull/13497
on the justification that some GC may need to be notified even for
immediate.
But the two currently available GCs don't, and there are plenty
of assumtions GCs don't ever... -
- * This reverts commit 623559faa3dd0927b4034a752226a30ae8821604.
* There is an issue with the jump in LIR, see https://github.com/ruby/ruby/pull/15542. -
07:59 PM Feature #21785 (Open): Add signed and unsigned LEB128 support to pack / unpack
- Hi,
I'd like to add signed and unsigned LEB128 support to the pack and unpack methods. LEB128 is a variable length encoding scheme for integers. You can read the wikipedia entry about it here: https://en.wikipedia.org/wiki/LEB128
... -
07:56 PM Revision eaa952b5 (git): YJIT: Print `Rc` strong and weak count on assert failure
- For <https://bugs.ruby-lang.org/issues/21716>, the panic is looking like
some sort of third party memory corruption, with YJIT taking the fall.
At the point of this assert, the assembler has dropped, so there's
nothing in YJIT's code oth... -
- For <https://bugs.ruby-lang.org/issues/21716>, the panic is looking like
some sort of third party memory corruption, with YJIT taking the fall.
At the point of this assert, the assembler has dropped, so there's
nothing in YJIT's code oth... -
- This reverts commit 2f151e76b5dc578026706b31f054d5caf5374b05.
The SP decrement (push) before the call do not match up with
the pops after the call, so registers were restored incorrectly.
Code from:
./miniruby --zjit-call-threshol... -
-
07:42 PM Misc #15487 (Closed): Clarify default gems maintanance policy
-
-
- Ugh, for some reason this was forgotten.
Any chance it can be updated/merged before 4.0, or is it too late already? Not a big issue, I'd guess, but would be good to wrap it up some time.
@byroot Can I be of any help? -
07:07 PM
Bug #19112 (Closed): Ractor garbage collection breaks/disables all active tracepoints
- Applied in changeset commit:git|4fb537b1ee28bb37dbe551ac65c279d436c756bc.
----------
Make tracepoints with set_trace_func or TracePoint.new ractor local (#15468)
Before this change, GC'ing any Ractor object caused you to lose all
enabl... - 07:06 PM Revision 4fb537b1 (git): Make tracepoints with set_trace_func or TracePoint.new ractor local (#15468)
- Before this change, GC'ing any Ractor object caused you to lose all
enabled tracepoints across all ractors (even main). Now tracepoints are
ractor-local and this doesn't happen. Internal events are still global.
Fixes [Bug #19112] -
- Applied in changeset commit:git|d209e6f1c0a93ad3ce1cc64dd165a6b67672614d.
----------
search_nonascii(): Replace UB pointer cast with memcpy
Casting a pointer to create an unaligned one is undefined behavior in C
standards. Use memcpy t... -
- Casting a pointer to create an unaligned one is undefined behavior in C
standards. Use memcpy to express the unaligned load instead to play by
the rules.
Practically, this yields the same binary output in many situations
while fixing th... -
-
04:47 PM Revision aab4f628 (git): Add the instance variable name and the module in Ractor::IsolationError (#15563)
-
- > However, since that functionality actually exists, I think #size has no choice but to return nil.
Respectfully, I disagree. I think it is much easier _and more useful_ to explain it along the lines of...
**It is infinite by impl... -
03:45 PM Bug #21780: Change the default size of Enumerator.produce back to infinity
- I understand zverok's feeling. In fact, I thought `Enumerator.produce` always returned an infinite Enumerator. I'm surprised it can be stopped by `StopIteration`. However, since that functionality actually exists, I think `#size` has no ...
-
- @mame My thinking (already outlined above) goes this way:
* while `Enumerator#size`, I believe, is not used extensively, it might in the future (this very change of `Enumerator.produce` call-sequence brings some attention to it)
* `Enu... -
- Yes, all bug fixes could be incompatibilities. What I am interested in is whether this incompatibility is actually causing a real problem, and if so, what kind of applications or libraries was affected.
-
02:54 PM Bug #21780: Change the default size of Enumerator.produce back to infinity
- The point of this discussion is that fixing the "bug" broke compatibility, and we are comparing the impacts. That "bug" can or should be properly fixed by changing the code to `Enumerator.produce(1, size: nil) { raise StopIteration }.la...
-
- With Ruby 3.4:
```ruby
Enumerator.produce(1) { raise StopIteration }.lazy.take(5).size #=> 5
Enumerator.produce(1) { raise StopIteration }.lazy.take(5).to_a.size #=> 1
```
I believe it is fair to call this behavior a bug. -
- mame (Yusuke Endoh) wrote in #note-16:
> What problem do you have if the size is `nil`?
Among others, `Enumerator.produce(1) { it+1 }.lazy.take(5)` now returns nil, which returned 5 previously.
https://github.com/ruby/ruby/blob/6b... -
- I couldn't find any cases where `Enumerator#size` returns `Float::INFINITY` for a finite-length Enumerator, except `Enumerator.produce` and when explicitly creating a fake-sized enumerator with `Enumerator.new(Float::INFINITY)`. Am I mis...
-
- knu (Akinori MUSHA) wrote in #note-14:
> I'm leaning toward doing these:
> ...
IMO, this seems like a great compromise! -
- I'm leaning toward doing these:
- Removing the Enumerator#to_set override that refuses to work against an infinite enumerator as a safeguard
- Reverting the default size of Enumerator.produce from nil to infinity
This way we can g... -
- zverok (Victor Shepelev) wrote in #note-7:
> 2) Somebody uses `Enumerator.produce` alongside other types of enumerators. In some branch of their code, they do `raise "Can't do this operation" if enum.size == Float::INFINITY`. The compati... -
- It's good points about compatibility, I agree losing the `size` on `.take` is not good.
IMO trying to detect "infinite loop" (in https://bugs.ruby-lang.org/issues/21654) is kinda pointless, it's the halting problem, for the vast majorit... -
- > This is where you are mistaken.
No, that's exactly what I've meant:
```ruby
# case 1:
e = Enumerator.produce(1, size: Float::INFINITY) {it>=3 ? raise(StopIteration) : it+1}
p e.to_set
# ruby 3.4: #<Set: {1, 2, 3}>
# master... -
- I may have misread your comment. Let me review again.
-
- zverok (Victor Shepelev) wrote in #note-7:
> TBH, I don't see the compatibility argument applied with any consistency here.
> ...
This is where you are mistaken.
- Ruby 3.4.7
```shellsession
% ruby -ve 'e=Enumerator.produce(1) {it>... -
- > However, the change that made Enumerator#to_set refuse to operate when the size returns infinity introduced a compatibility issue
TBH, I don't see the compatibility argument applied with any consistency here.
Let's imagine severa... -
- The argument that Enumerator.produce is infinite by nature is certainly valid. However, the change that made Enumerator#to_set refuse to operate when the size returns infinity introduced a compatibility issue: it breaks existing code th...
-
-
- I have the following hypothesis: When people want column information, what they really need is an AST. #21005 was precisely such an example.
In some cases, column information alone may suffice, but obtaining the AST inevitably yields co... -
- > Regarding to power_assert
So it's great power_assert has much more advanced parsing capabilities, but I for one would use the column information for things like Rails' `assert_difference -> { User.count }, +1`, as it would be quite ... -
- > And also to be able to use `Prism.node_for` one of course needs `Proc#source_location`.
Huh? `Proc` already has its `node_id` (via its ISeq), so `Prism.node_for` is implementable without `source_location`.
In fact, by doing the follo... -
- Yes, it'd be useful for e.g. `assert_raise(SomeException) { ... }` in test-unit, `expect { ... }.to` in RSpec, `-> { ... }.should raise_error(...)` in MSpec, etc.
For example `-> { 13 - "10" }.should raise_error(ArgumentError)` in MSp... -
- Isn't `Proc.source_location` very useful for things akin to `power_assert`?
If it was easy to extract a proc's source code I'd certainly integrate it in testing framework to improve failure rendering. -
- mame (Yusuke Endoh) wrote in #note-13:
> I agree. But then, when would `Proc#source_location` be useful?
There are cases where it's useful to show just the block, i.e. `{ block body }`.
And also to be able to use `Prism.node_for` on... -
- Eregon (Benoit Daloze) wrote in #note-12:
> And I suspect many of these cases would likely use `Prism` and specifically [Prism.node_for](https://github.com/ruby/prism/pull/3808) to get more information, such as showing the method call t... -
- byroot (Jean Boussier) wrote in #note-10:
> Maybe I'm totally off, but I expect this data to be used to extract the source code, e.g show a snippet of code in an error message, or something akin to that, hence byte offsets seem actually... -
- I made a PR to re-add `{Method,UnboundMethod,Proc}#source_location` and fix all known issues: https://github.com/ruby/ruby/pull/15580
@matz Would it be OK to merge it? 🙏
For context I opened this issue because I was surprised at the sem... -
- Maybe I'm totally off, but I expect this data to be used to extract the source code, e.g show a snippet of code in an error message, or something akin to that, hence byte offsets seem actually more convenient? (and performant).
But ye... -
- matz (Yukihiro Matsumoto) wrote in #note-8:
> I'd like to cancel `source_location` to have column information in 4.0, due to this concern.
> ...
Thank you for the quick reply, I think that would be the worst outcome though, https://bug... -
02:46 AM Bug #21783: {Method,UnboundMethod,Proc}#source_location returns columns in bytes and not in characters
- I'd like to cancel `source_location` to have column information in 4.0, due to this concern. In my personal opinion, I am leaning toward byte index, though.
Matz.
- 01:23 PM Revision 6b35f074 (git): Box: [DOC] Add RUBY_BOX in Environment
-
- knu (Akinori MUSHA) wrote in #note-17:
> Adding Range#to\_set and Enumerator#to\_set overrides that perform size checks would help users avoid most common pitfalls.
I think it's actually very rare to call `.to_set` on something else ... -
- Right, I thought about that too but forgot to write it.
Several gems have already adapted to this feature and their functionality and/or tests might rely on having the extra information, so it's definitely not great to remove this last ... -
11:26 AM Feature #6012: Proc#source_location also return the column
- And RSpec were also adjusted to the new behavior AFAICT:
https://github.com/rspec/rspec/pull/282/commits/1c20fa80772ca7a1ed0512056ce7cd6a94f8e68d -
- Just FTR, not sure how the revert is supposed to look like, but at least Pry was adjusted for the new behavior:
https://github.com/pry/pry/pull/2357
It does not benefit from columns, but I suspect the revert would somehow influence... -
- mame (Yusuke Endoh) wrote in #note-32:
> I record the summary of the discussion regarding Matz's decision to revert this for now. The decision was based on a combination of several reasons:
Thank you for documenting that.
It feels t... -
- I record the summary of the discussion regarding Matz's decision to revert this for now. The decision was based on a combination of several reasons:
* Unclear use cases
* The description of #6012 lacked a use case.
* The use cas... -
- Although last minute, we will cancel this feature in 4.0 because of design ambiguities such as whether to return column positions in bytes or characters as in #21783.
Matz.
- 10:52 AM Revision 2b1a9afb (git): Fix: Do not pass negative timeout to Addrinfo#connect_internal (#15578)
- This change fixes a bug where, with `Socket.tcp`’s `fast_fallback option` disabled, specifying `open_timeout` could unintentionally pass a negative value to `Addrinfo#connect_internal`, `causing an ArgumentError`.
```
❯ ruby -rsocket -e... -
- Since it now live in the EC.
-
- Mutexes spend a significant amount of time in `rb_fiber_serial`
because it can't be inlined (except with LTO).
The fiber struct is opaque the so function can't be defined as inlineable.
Ideally the while fiber struct would not be opaque... -
07:53 AM Revision 85b40c5e (git): Box: fix the class name in tests
-
- Also remove a stale TODO.
-
-
-
-
- and columns"
This reverts commit 073c4e1cc712064e626914fa4a5a8061f903a637.
https://bugs.ruby-lang.org/issues/6012#note-31
> we will cancel this feature in 4.0 because of design ambiguities
> ...
[#21783]: https://bugs.ruby-lang.org/iss... -
06:28 AM Revision 5b0fefef (git): [ruby/rubygems] Added assertion for Windows and nmake
- https://github.com/ruby/rubygems/commit/be5c4e27d9
-
- https://github.com/ruby/rubygems/commit/09e6031a11
- 04:21 AM Revision 080bf30c (git): [ruby/rubygems] Allow to specify the number of `make` jobs when installing gems:
- - Added a new `-j` option to `gem install` and `gem update`.
This option allows to specify the number of jobs we pass to `make`
when compiling gem with native extensions.
By default its the number of processors, but users may want... - 04:21 AM Revision 9f593156 (git): [ruby/rubygems] Pass down value of `BUNDLE_JOBS` to RubyGems before compiling:
- - ### Problem
Since https://github.com/ruby/rubygems/pull/9131, we are now
compiling make rules simultaneously. The number of jobs
is equal to the number of processors.
This may be problematic for some users as they want to cont... -
04:05 AM Bug #21266 (Closed): YJIT GC safety crash with proc objects as block argument
- Applied in changeset commit:git|9168cad4d63a5d281d443bde4edea6be213b0b25.
----------
YJIT: Bail out if proc would be stored above stack top
Fixes [Bug #21266]. -
- Fixes [Bug #21266].
- 03:56 AM Revision f3b9509b (git): [ruby/rubygems] Fix quote handling in mise format ruby version parsing
- The previous regex didn't properly match quoted strings
it would capture the opening quote as part of the version
if quotes were mismatched.
This change properly parses double-quoted, single-quoted,
and unquoted version strings separatel... - 03:56 AM Revision 3b50f4ba (git): [ruby/rubygems] Support single quotes in mise format ruby version
- https://github.com/ruby/rubygems/commit/a7d7ab39dd
-
- This can happen with documentation updates and we don't want
those to trip on ZJIT tests.
Redact the whole name since names like "_bi342" aren't that helpful
anyways. -
- The column information from `source_location` will be canceled in 4.0. Let us discuss for the future. I think it's a matter of taste. I don't really care much.
Matz.
-
- https://github.com/ruby/rubygems/commit/a091e3fd10
- 12:56 AM Revision 38d67986 (git): Bump actions/cache in /.github/actions/setup/directories
- Bumps [actions/cache](https://github.com/actions/cache) from 5.0.0 to 5.0.1.
- [Release notes](https://github.com/actions/cache/releases)
- [Changelog](https://github.com/actions/cache/blob/main/RELEASES.md)
- [Commits](https://github.co... -
12:25 AM Revision b6d4562e (git): [DOC] Remove copyright from Set class docs
- 12:16 AM Revision ff1b8ffa (git): [ruby/rubygems] Tweak the Bundler's "X gems now installed message":
- - Fix https://github.com/ruby/rubygems/pull/9188
- This message is a bit misleading because it always outputs one extra
specs, which is Bundler itself.
This is now fixed when the message is about to be output.
https://github.com/ru... - 12:15 AM Revision f88e7970 (git): [ruby/rubygems] Allow bundle pristine to work for git gems in the same repo:
- - Fix https://github.com/ruby/rubygems/pull/9186
- ### Problem
Running `bundle pristine` in a Gemfile where there is many git gem
pointing to the same repository will result in a error
"Another git process seems to be running in t... -
12:10 AM Revision 98cac1a7 (git): Point people to redmine on ZJIT docs (#15499)
- Fix https://github.com/Shopify/ruby/issues/900
12/15/2025
-
- * Fixes https://github.com/ruby/psych/issues/760
https://github.com/ruby/psych/commit/952008c898 - 10:48 PM Revision b3f0fb56 (git): [ruby/psych] Replace C extension with Data#initialize bind_call
- https://github.com/ruby/psych/commit/6a826693ba
-
10:28 PM Revision cfd41cbf (git): [DOC] Harmonize #-@ methods
-
-
- 10:07 PM Revision 98ab418f (git): Revert "Fix Socket.tcp cleanup after Thread#kill (#15131)" (#15565)
- This reverts commit 3038286a4bf7832f1c42c8cc9774ee6ff19876fc.
The following CI failure scared me:
https://github.com/ruby/ruby/actions/runs/20241051861/job/58108997049
```
1) Timeout:
TestResolvDNS#test_multiple_servers_with_timeou... -
- I made a PR to fix this: https://github.com/ruby/ruby/pull/15568
Looking at https://github.com/ruby/ruby/pull/9872/files it's clear in Prism it was just done to be compatible with parse.y.
And in parse.y it's probably not intentional... -
- I too, would expect the whole expression to be covered. This would be also consistent with `Method#source_location` which includes `def` (and not starts from the argument, or from the beginning of the body).
-
- Interesting.
The arguments (parameters) between `->` and `{` are AFAIK integral part of the stabby lambda syntax, so I think it should be either:
* All cases start at the `{`:
* consistent with non-stabby-lambda Procs
* but does ... -
03:35 PM Bug #21784: Proc#source_location start column seems strange for -> {}
- Except the problematic case `-> { }`, start position consistently excludes whitespaces just after `->`.
So I think `-> { }`'s source location should be `{ }`.
~~~ruby
-> ( ) { }.source_location # ["(irb)", 1, 3, 1, 10]
^^^^^^^
->... -
- ```
$ ruby -e 'p -> { a }.source_location'
["-e", 1, 4, 1, 10]
```
So it considers the stabby lambda to be at:
```ruby
p -> { a }.source_location
^^^^^^
```
Note it starts at a whitespace after the `->`.
I think it should ... -
08:27 PM Bug #21783: {Method,UnboundMethod,Proc}#source_location returns columns in bytes and not in characters
- Honestly if we're interpreting column as something visual like you're implying, we're also going to run into issues with grapheme clusters and east asian width and all the other implications for whatever "character" actually means. I thi...
-
- From https://bugs.ruby-lang.org/issues/6012#note-25 @matz said adding column was OK, but not byte offsets.
I'm not sure what were his reasons, but maybe it's that byte offsets are too low-level for `source_location`?
If so, I would thi... -
- Updating the docs is one solution, so at least it's consistent between docs and behavior.
I think as a Ruby-facing API it's weird that it operates in terms of bytes (and `source_location` does not have a `byte` prefix to indicate that... -
- I think this is a documentation issue, as both parsers/compilers operate in terms of bytes. Changing this to characters would likely be a noticeable difference in speed, and quite a bit of code change. (Either both parsers/compilers woul...
-
- The documentation says:
```
= Proc.source_location
(from ruby core)
------------------------------------------------------------------------
prc.source_location -> [String, Integer, Integer, Integer, Integer]
---------------... - 05:16 PM Revision 74b24e09 (git): Update default gems list at adf676c530b5da2822aa3b03ee43e5 [ci skip]
-
05:15 PM Revision adf676c5 (git): [ruby/erb] Version 6.0.1
- https://github.com/ruby/erb/commit/bbde68fcd5
-
05:14 PM Revision fdd8bdea (git): [ruby/erb] Freeze ERB::Compiler::TrimScanner::ERB_STAG
- (https://github.com/ruby/erb/pull/100)
For Ractor compatibility.
https://github.com/ruby/erb/commit/43f0876595 -
- Run like so:
$ ../tool/zjit_iongraph.rb ../build-dev/miniruby --zjit-call-threshold=2 tmp/ghbug.rb
false
false
tmp/ghbug.rb:3:in 'Object#doit': this shouldnt ever be nil (RuntimeError)
from tmp/ghbug.rb:10:in... - 04:48 PM Revision 3038286a (git): Fix Socket.tcp cleanup after Thread#kill (#15131)
- Socket.tcp launches ruby threads to resolve hostnames, and those threads
communicate through a queue implemented with `IO.pipe`. When the thread
that called `Socket.tcp` is killed, the resolver threads still try to
communicate through th... -
04:28 PM Revision 6b63b0cb (git): [DOC] Update Set#inspect description in NEWS
-
- Otherwise Array#detect is Enumerable#detect while Array#find uses a
different more performant implementation.
[Feature #21678] -
- Is this an issue where you could have defined `Enumerable#find` and now it won't have overwritten `Array#find`? Should I revert that part of this patch?
-
- @etienne Yes, that seems clearly a good fix, could you do it?
(I personally don't like `detect` but we should keep aliases consistent in performance and behavior) -
- Should we alias `detect` to also use that Array-specific faster implementation?
```console
>> [].method :detect
=> #<Method: Array(Enumerable)#detect(*)>
> ...
=> #<Method: Array#find(*)>
```
Edit: https://github.com/ruby/ruby/pull/15558 -
- https://github.com/ruby/prism/commit/609c80c91e
-
- Right, I think Regexp interpolation should be closer to String interpolation, currently it's its own separate thing with rather weird rules.
It reminds me of some other issues related to Regexp interpolation like #20407 and linked issues. -
- Right, it's too late for 4.0, OK.
> Frankly, I don't think freezing the Regexp objects is worth the effort/trouble.
I already did the implementation effort so that part is fine.
I think it's worth it for consistency and clean sema... -
- When you have a regular expression that has a named capture that
has an escape sequence in the named capture, and that escape
sequence is a unicode escape sequence with an invalid surrogate
pair, the error was attached to the owned strin... -
-
- It's a bit late to mention this, but I find this new behavior quite strange.
Range is used to represent both discrete sequences and continuous ranges, and when performing iterative operations, the begin value distinguishes between them.... -
- I have discovered that this issue reproduces when Gemfile contains a gem referred by `path:` or `git:`.
Such Gemfiles are pretty much common in real-world applications including Rails, which use in-house Gems and libraries, like this:
... -
- Patch submitted: https://github.com/ruby/ruby/pull/15556
-
- Ractor::IsolationError is raised when non-main Ractors attempt access to non-shareable constants. The message contains the path to the constant which triggered the violation.
However, the path is incorrect when the constant was resolved... - 09:51 AM Revision 35209cae (git): Update default gems list at f0793731853c0e130f798e9dc5c736 [ci skip]
-
09:50 AM Revision f0793731 (git): [ruby/openssl] Ruby/OpenSSL 4.0.0
- https://github.com/ruby/openssl/commit/5af1edab18
-
- https://github.com/ruby/openssl/commit/b0de8ba9bd
-
- https://github.com/ruby/openssl/commit/695126f582
-
08:44 AM Bug #21715: Miscompilation on x86-64-v2 due to undefined behavior in search_nonascii in string.c
- I investigated further and I understand Alan's patch is the best at this time.
auto vectorization for this kind of loop is implemented only by GCC 15, it's too early.
@Alan, could you commit it? -
-
-
-
-
-
- * Make invariant `root_box` an array consist of only `root_box_data`.
* Remove the unnecessary initializer list that is just overwritten in
`initialize_root_box()` and missing `classext_cow_classes`.
* Shrink the scope using another lo... -
- This contains various improvements in tests for openssl integration:
- Remove DHE parameters from test servers. OpenSSL is almost always
compiled with ECC support nowadays and will prefer ECDHE over DHE.
- Remove an outdated omi... - 03:11 AM Revision a2dc4d7f (git): Bump actions/upload-artifact from 5.0.0 to 6.0.0
- Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 5.0.0 to 6.0.0.
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](https://github.com/actions/upload-artifact/compare/v5...v... - 02:33 AM Revision 76f30030 (git): Bump msys2/setup-msys2 from 2.29.0 to 2.30.0
- Bumps [msys2/setup-msys2](https://github.com/msys2/setup-msys2) from 2.29.0 to 2.30.0.
- [Release notes](https://github.com/msys2/setup-msys2/releases)
- [Changelog](https://github.com/msys2/setup-msys2/blob/main/CHANGELOG.md)
- [Commits... - 02:31 AM Revision dd25fdcd (git): Bump actions/cache from 5.0.0 to 5.0.1
- Bumps [actions/cache](https://github.com/actions/cache) from 5.0.0 to 5.0.1.
- [Release notes](https://github.com/actions/cache/releases)
- [Changelog](https://github.com/actions/cache/blob/main/RELEASES.md)
- [Commits](https://github.co... -
- Reline has been moved to a bundled gem, so we don't need the docs anymore.
-
01:15 AM Revision 58940377 (git): [ruby/rubygems] Write gem files atomically
- This change updates `write_binary` to use a new class,
`AtomicFileWriter.open` to write the gem's files. This implementation
is borrowed from Active Support's [`atomic_write`](https://github.com/rails/rails/blob/main/activesupport/lib/ac...
12/14/2025
-
-
09:11 PM Revision 490a03ba (git): [ruby/prism] Fix `sexp_processor` gem reference
- It's https://rubygems.org/gems/sexp_processor, not https://rubygems.org/gems/sexp
https://github.com/ruby/prism/commit/b8a00a5f15 -
- There are, by the way, other effects of the current default that are, even if minor, still annoying:
```ruby
Enumerator.produce(1, &:succ).lazy.take(6).size
# Ruby 3.4: => 6 -- which is correct and useful
# Ruby 4.0: => nil -- ... -
- > I think `Enumerator#size` should only be non-nil when it is known to be the exact size.
> ...
I would argue that it is _known_ to be infinite: that's how `produce` works: loops infinitely, unless explicitly stopped by an _exception_, t... -
- Actually for Enumerator.new, it's trivial to not be infinite and does not even need StopIteration, e.g.:
```ruby
Enumerator.new { |y| y << 1 }.count # => 1
```
So I guess you meant to use `Enumerator.produce` instead in your exampl... -
- I disagree on this one, as written on https://bugs.ruby-lang.org/issues/21701#note-3
I think `Enumerator#size` should only be non-nil when it is known to be the exact size.
In this case it is not known if it is infinite, so returning `F... -
- In #21701 a new argument `size:` was introduced, and its default value is `nil` (unknown).
While I support the new argument, I'd argue that the default should be `Float::INFINITY`.
**Reasoning:** By _design_, `Enumerator.produce` ... -
-
-
- In https://bugs.ruby-lang.org/issues/10017, `fetch_values` was introduced for `Hash`. `ENV` is hash-like but is missing the method. Here are the differences between hash and env right now:
```rb
irb(main):006> Hash.instance_methods -... -
- https://github.com/ruby/prism/commit/65595d6c2c
-
- https://github.com/ruby/prism/commit/fc150b1588
- 04:42 PM Revision 4a84fa1b (git): [ruby/prism] Define RubyParser::SyntaxError directly and drop require for ruby_parser.
- Had to add a require of sexp since that came in indirectly via ruby_parser.
https://github.com/ruby/prism/commit/df677c324f -
- Here is my PR to add `Prism.node_for` as first suggested by @mame in https://bugs.ruby-lang.org/issues/21005#note-5:
https://github.com/ruby/prism/pull/3808 -
- When d2i_PKCS7_bio() and PEM_read_bio_PKCS7() fail to decode the input,
OpenSSL::PKCS7.new currently raises ArgumentError. The usual practice
in ruby/openssl where an error originates from the underlying OpenSSL
library is to raise OpenS... -
- Mention the underlying OpenSSL function. Add a note about the unreliable
comparison when called on an incomplete object.
Fixes https://github.com/ruby/openssl/issues/844
https://github.com/ruby/openssl/commit/736af5b3c7 -
- @jeremyevans0 Thank you for the clarification. I'll try to consider adjusting the docs to reflect the design.
-
02:46 AM Feature #21389 (Closed): Simplify Set#inspect output
- zverok (Victor Shepelev) wrote in #note-5:
> Just noticed that on the latest `master` subclasses `#inspect` didn't change (unlike what @matz suggests here: https://bugs.ruby-lang.org/issues/21389#note-3):
> ...
Yes, this is deliberate ... -
- @mame Thank you for the clarification!
-
- This was pointed by @ko1 at the dev meeting in the discussion of #21049, and we agreed not to implement it. Implementing it would require the parser to accept more detailed outer context, and that effort was not worth. Sorry it wasn't cl...
-
- * [Bug #21780] Change the default size of `Enumerator.produce` back to infinity (zverok)
* While the new argument is a good addition to the API, I argue that `Float::INFINITY` would be a more friendly default, corresponding to the mos... -
- 02:59 AM Revision ab95abd4 (git): Update default gems list at 9e22037eb50d7abe6385901c9b0293 [ci skip]
-
- https://github.com/ruby/io-console/commit/fbc7e1f31f
-
- Fix up ruby/ruby#15414, 29c29c2b7e972359ab83038c5dc27a7e53ae65c7
-
- The `--zjit-stats-quiet` and `--zjit-stats=quiet` options differ.
The latter option, `=quiet`, does print stats to the file "quiet", but
does not suppress output like yjit option `--yjit-stats=quiet`.
Fix up ruby/ruby#15414, 29c29c2b7e9... -
- Applied in changeset commit:git|bc2a8a002a6c41fc1b28e02e15e2fb2b72d1b66e.
----------
[Bug #21779] Uniquify `InitVM` functions as well as `Init`
Avoid possible name conflict when `--with-static-linked-ext`. -
- Reported at https://github.com/ruby/io-console/pull/105.
This is caused by `InitVM_console` exposed from the statically linked io/console is called from the loaded io-console gem.
The static `InitVM_console` initializes the ractor lo... -
- Fix ruby/io-console#105.
-
- Avoid possible name conflict when `--with-static-linked-ext`.
-
-
12/13/2025
-
-
-
- Oh, that's a very interesting trick @tompng
-
- Maybe this is off-topic, but I think this stack that internally use queue can perform closer to Thread::Queue.
~~~ruby
def pop(timeout: nil)
if @dummy_resource_queue.deq(timeout:)
@real_resource_array.pop
end
end
def <<(... -
- I'm +1 on making Monitor core.
(FWIW TruffleRuby already uses `Primitive` to define Monitor as basically core: https://github.com/truffleruby/truffleruby/blob/master/lib/mri/monitor.rb) -
- So, if Monitor was made a core class, it could benefit from some optimizations and be almost as fast as Mutex: https://github.com/ruby/ruby/pull/15538
```
ruby 4.0.0dev (2025-12-13T06:49:18Z core-monitor 6fabf389fd) +YJIT +PRISM [arm... -
- Yeah, sure. I compiled them in this gist: https://gist.github.com/Earlopain/79d49df05cca5b0c8c6cefc6f3284a4b. I missed one vendored file, so there are only 6 distinct cases.
In `ruby/prism` I added a script for this. You call it like ... -
- Earlopain (Earlopain _) wrote in #note-12:
> I ran this change over code available on rubygems and it only impacts 37 files which go syntax invalid with this. If I take out duplicate vendorered files, it is down to just 7.
If it's no... -
- Yeah, it seems like I missed this case but @nobu already took care of it in his PR. I left a comment about something, however since this is in ruby/ruby now I can't actually push any changes anymore myself.
I ran this change over code... -
- Just noticed that on the latest `master` subclasses `#inspect` didn't change (unlike what @matz suggests here: https://bugs.ruby-lang.org/issues/21389#note-3):
```ruby
class MySet < Set
end
p Set[1, 2, 3] #=> Set[1, 2, 3]
p My... -
- This works:
```ruby
proc { |x| binding.eval('x') }.call(1) #=> 1
```
This doesn't (neither with numbered parameters, nor with `it`):
```ruby
proc { _1; binding.eval('_1') }.call(1) # undefined local variable or method '_1' for ... -
- An OpenSSL function sometimes puts more than one error entry into the
thread-local OpenSSL error queue. Currently, we use the highest-level
entry for generating the exception message and discard the rest.
Let ossl_make_error() capture a... -
- Propagate the value given with `--with-git` configure option to
tool/lib/vcs.rb. -
02:33 PM
Feature #21701 (Closed): Enumerator.produce accepts an optional `size` keyword argument
- Applied in changeset commit:git|79a6ec74831cc47d022b86dfabe3c774eaaf91ca.
----------
Enumerator.produce accepts an optional `size` keyword argument
When not specified, the size is unknown (`nil`). Previously, the size was always `Floa... - 02:33 PM Revision 79a6ec74 (git): Enumerator.produce accepts an optional `size` keyword argument
- When not specified, the size is unknown (`nil`). Previously, the size was always `Float::INFINITY` and not specifiable.
[Feature #21701] - 02:29 PM Revision 0561eb94 (git): [ruby/prism] Prevent an infinite loop parsing a capture name
- Fixes https://github.com/ruby/prism/pull/3729.
https://github.com/ruby/prism/commit/6e5347803c -
- https://github.com/ruby/io-console/commit/379e7c17ed
-
12:37 PM Revision 3a4ad76f (git): [ruby/io-console] console_cursor_pos respects scroll position on windows
- https://github.com/ruby/io-console/commit/ae33785820
-
- https://github.com/ruby/io-console/commit/d2a6c69697
-
- On windows, IO.console.winsize= now respects the current view area and screen buffer size.
https://github.com/ruby/io-console/commit/817aa65ea3 -
- Thanks @k0kubun. Seems like I can, yeah. I'll do that in the future where I think it makes sense.
-
- https://github.com/ruby/io-console/commit/df444b93f1
-
04:52 AM Revision 71dd2725 (git): Remove useless rb_check_arity() calls
-
- Credits to @rwstauner for noticing this issue in GH-15533.
-
- This reverts commit 1eb10ca3cb6cff98bb8c0946ed905921586c7d52.
This should have been fixed by https://github.com/ruby/ruby/pull/15536. -
- Avoid garbage reads from locals in eval.
Before the fix the test fails with
<"[\"x\", \"x\", \"x\", \"x\"]"> expected but was
<"[\"x\", \"x\", \"x\", \"x286326928\"]">.
12/12/2025
-
-
- Credits to @rwstauner for noticing this issue in GH-15533.
-
- Previously, the chain_depth>0 version of setlocal blocks did not
update the type of the local variable in the context. This can leave
the context with stale type information and trigger or lead to miscompilation.
To trigger the issue, Y... -
11:29 PM Revision 2f151e76 (git): ZJIT: Allow ccalls above 7 arguments (#15312)
- ZJIT: Add stack support for CCalls
-
- Thanks for sharing the logs -- I was able to track down the issue using it.
Backport for 3.4 is pending at https://github.com/ruby/ruby/pull/15535 -
- Applied in changeset commit:git|2884f53519c4b86072d5fc3b41a71cee697af8ba.
----------
YJIT: Add missing local variable type update for fallback setlocal blocks
Previously, the chain_depth>0 version of setlocal blocks did not
update the ... -
10:30 PM Bug #21696 (Closed): Performance degradation for long running processes in Ruby 4.0.0-preview2
-
10:03 PM Revision 176e384b (git): Cache filesystem_encindex
-
- When debugging the fstring table, I found "UTF-8" to be the most common
interned strings in many benchmarks.
We have a fixed, limited number of these strings, so we might as well permanently
cache their fstrings. -
- Previously, the chain_depth>0 version of setlocal blocks did not
update the type of the local variable in the context. This can leave
the context with stale type information and trigger panics like in
[Bug #21772] or lead to miscompilati... - 07:47 PM Revision 3add3db7 (git): Fewer calls to `GET_EC()` and `GET_THREAD()` (#15506)
- The changes are to `io.c` and `thread.c`.
I changed the API of 2 exported thread functions from `internal/thread.h` that
didn't look like they had any use in C extensions:
* rb_thread_wait_for_single_fd
* rb_thread_io_wait
I didn't cha... -
- Applied in changeset commit:git|5903ed7ba9ca60546aa0dd97e92b3d381b7918d3.
----------
Prevent ifunc procs from being made shareable
[Bug #21775] -
- [Bug #21775]
- 07:24 PM Revision 7909ce2a (git): move th->event_serial to rb_thread_sched_item (#15500)
-
- Applied in changeset commit:git|6147b695870ce82ee3ad5305ce095b63889b8d9d.
----------
Array#rfind
Implement Array#rfind, which is the same as find except from the
other side of the Array. Also implemented Array#find (as opposed to
the g... -
- Implement Array#rfind, which is the same as find except from the
other side of the Array. Also implemented Array#find (as opposed to
the generic one on Enumerable because it is significantly faster
and to keep the implementations togethe... - 06:26 PM Revision 4f900e3c (git): ZJIT: Only optimize `[]` and `[]=` for exact Hash, not Hash subclasses
- 06:26 PM Revision 309d6ef9 (git): ZJIT: Inline `Hash#[]=`
-
- It's just a nicety (they fit fine as CUInt32) but this makes printing
look nicer in real execution and also in tests (helps with #15489).
Co-authored-by: Randy Stauner <randy@r4s6.net> - 05:52 PM Revision bb4a6f39 (git): [ruby/prism] Fixed Prism::Translation::RubyParser's comment processing
- Tests were failing in Flay under Prism.
https://github.com/ruby/prism/commit/af9b3640a8 -
- I'm not a big fan the pragma route. Trying to get good codegen out of UB triggering C code is inherently a whack-a-mole game with compiler brands and even options of the same brand. With non compliant C code the incantation to patch over...
-
- Previously, if an argument splat and keywords are provided by
the caller, it did not check whether the method/proc accepted
keywords before avoiding the allocation. This is incorrect,
because if the method/proc does not accept keywords, ... -
-
- From: https://github.com/ruby/ruby/pull/15529
> The timer thread checks the running threads to provide timeslice every 10ms and finds the thread (th) performing blocking operation.
Isn't 10ms a massive amount of time in that contex... -
- I'm with John that I think this is very promising but should not be introduced in ruby 4.0. It needs more testing to make sure it's working correctly and it needs more benchmarks for more realistic workloads. We can't do that in 2 weeks....
-
02:01 PM Bug #21685: Unnecessary context-switching, especially bad on multi-core machines.
- https://github.com/ruby/ruby/pull/15529
It seems enough small, so I want to introduce it to 4.0. -
- # The next dev meeting
**Date: 2026/01/14 13:00-17:00** (JST)
Log: *TBD*
- Dev meeting *IS NOT* a decision-making place. All decisions should be done at the bug tracker.
- Dev meeting is a place we can ask Matz, nobu, nurse and o... -
-
- > I feel Unix command names are too short for usual programs.
At the same time:
* they are known at least to most of the console-using programmers, so this is a "dictionary" many of us familiar with
* FileUtils have a practice of al... -
12:29 AM Feature #21766: Pathname + FileUtils making sweet music together
- I feel Unix command names are too short for usual programs.
So, I'm negative. -
- Applied in changeset commit:git|fa7cddc969f1eccbb377cfc752bbf82ee2887dde.
----------
Add Binding#implicit_parameters, etc.
This changeset introduces:
* `Binding#implicit_parameters`
* `Binding#implicit_parameter_get`
* `Binding#implic... -
- [Bug #21049]
-
-
- [Bug #21049]
-
- This changeset introduces:
* `Binding#implicit_parameters`
* `Binding#implicit_parameter_get`
* `Binding#implicit_parameter_defined?`
[Bug #21049] -
10:53 AM Misc #20774: Remove remaining locale dependent code from Windows port
- @nobu IMHO ruby-4.0 is a good chance to change these remaining locale strings to UTF-8. Windows stdin console input is still in [locale encoding](https://github.com/ruby/ruby/blob/d428d086c23219090d68eb2d027498c6ea999b89/io.c#L13557), th...
-
- `thread_sched_to_waiting_common0` is no longer needed.
- 09:12 AM Revision e2fe0aae (git): Avoid race condition in `test_without_handle_interrupt_signal_works`. (#15504)
-
- So I did a number of other patches to squeeze some more performance out of `Monitor#synchronize`
```
ruby 4.0.0dev (2025-12-12T09:08:05Z master ff831eb057) +YJIT +PRISM [arm64-darwin25]
Warming up -----------------------------------... -
- Saves one more call to GET_EC()
-
- It's more consistent with Mutex, but also the `#wait` method
benefits from receiving the execution context instead of having
to call `GET_EC`. - 08:55 AM Revision cf97a14c (git): Bump actions/cache from 4.3.0 to 5.0.0
- Bumps [actions/cache](https://github.com/actions/cache) from 4.3.0 to 5.0.0.
- [Release notes](https://github.com/actions/cache/releases)
- [Changelog](https://github.com/actions/cache/blob/main/RELEASES.md)
- [Commits](https://github.co... - 08:55 AM Revision 0022a878 (git): Bump actions/cache in /.github/actions/setup/directories
- Bumps [actions/cache](https://github.com/actions/cache) from 4.3.0 to 5.0.0.
- [Release notes](https://github.com/actions/cache/releases)
- [Changelog](https://github.com/actions/cache/blob/main/RELEASES.md)
- [Commits](https://github.co... -
- https://github.com/ruby/timeout/commit/4ae8631acf
- 06:55 AM Revision d9cc7621 (git): Update bundled gems list as of 2025-12-12
-
- Applied in changeset commit:git|1794cfe12fe61dedebadead542927f9fef4104eb.
----------
Binding#local_variable_defined? raises a NameError for numbered params.
[Bug #21776] -
-
- I noticed that `binding.local_variable_defined?(:@x)` raises `wrong local variable name '@x' for #<Binding:0x00007cd1e95a94d8> (NameError)`.
Maybe `binding.local_variable_defined?(:_1)` should raise a similar error because `_1` is not a... -
- Applied in changeset commit:git|04494d9e4064a57accc8309da9b35754a3d24973.
----------
`Binding#local_variable_defined?` must not handle numbered parameters
[Bug #21776] -
- Thanks! https://github.com/ruby/ruby/pull/15519 will fix the issue.
-
- instead of using `rb_native_mutex_lock` directly.
-
- [Bug #21776]
-
- Since support for Windows 9x was dropped over a decade ago.
-
- `dwMajorVersion` alone has no meaning since Windows 7. Use API in
VersionHelper.h instead. -
- It has been fixed at Windows 8, and we already have dropped the
support Windows 8 and olders. -
- [Bug #21776]
-
- and PREVIOUS_RELEASE_TAG for any .0 releases
-
-
- to share it with tool/releng/update-www-meta.rb and another place I'm
going to modify next. -
- ```
$ tool/format-release ../www.ruby-lang.org 4.0.0-preview2 .
/opt/rubies/3.4.6/lib/ruby/3.4.0/psych/class_loader.rb:99:in 'Psych::ClassLoader::Restricted#find': Tried to load unspecified class: Date (Psych::DisallowedClass)
fr... -
-
- ```
dir | ruby -e “puts STDIN.read.force_encoding('locale')”
```
I think they should use `-E`.
> On Windows, the C runtime locale for each process is currently inconsistent, so changing the locale encoding for this purpose does n... -
- Here's the list of all changes I made for this ticket:
* `tool/merger.rb tag`
* https://github.com/ruby/ruby/commit/0564214a00450371527c7bd69fc13618e5f25f30
* `tool/merger.rb remove_tag`
* https://github.com/ruby/ruby/commit/aff... -
-
- This is a refactoring change, which should have no impact on behaviors.
Now, if patchlevel is true, tag is empty. So `if patchlevel` always does
nothing. Given that prerelease is false and tag is not nil, removing
`if patchlevel` should... -
- It just seems like a completely different input, so it makes more sense
to me to have it as a separate case.
Also, we don't need to support Ruby 2.0 or older. -
- to make it a bit more consistent with the other branch
-
- We no longer make patchlevel releases.
-
- Just copied format-release fixes in
a973526c050fec044ffd7ceeba0ac8e8a1fed299 to this file.
12/11/2025
-
-
- The benchmark from #20816 gets a little more pure results as it only deals with a fast syscall in a loop without Queue communication. I adjusted it slightly.
``` ruby
THREADS = (ARGV.first || 10).to_i # choose thread count from CLI
... -
-
- All those PRs have been merged, so I did (3) as well.
-
-
- I filed PRs for (2).
* https://github.com/ruby/www.ruby-lang.org/pull/3689
* https://github.com/ruby/rbs/pull/2743
* https://github.com/docker-library/ruby/pull/522
I'll wait a little before doing (3). -
- I finished (1).
```
$ git checkout v4_0_0_preview2
HEAD is now at 4fa6e9938c strnlen is not used now
$ git tag v4.0.0-preview2
$ git push origin v4.0.0-preview2
Total 0 (delta 0), reused 0 (delta 0), pack-reused 0 (from 0)
remote: To ... -
- ## Changes
As per [Misc #21769], we'll rename the `v4_0_0_preview2` tag to `v4.0.0-preview2` on ruby.git.
## Operations
To minimize its potential impact, I'll work on it in the following order:
1. Push a new tag `v4.0.0-previ... -
- Thank you! I have also run the test over 10 times to make sure there are no failures with `RUBY_MN_THREADS=1`.
-
- This has most likely been fixed by https://github.com/ruby/ruby/pull/15392. I can't reproduce the failure on master and the fix was targeting a race condition in the thread scheduler in MN thread mode.
-
- Updated `tool/merge.rb tag` and `tool/format-release` to support the new git tag format at https://github.com/ruby/ruby/pull/15508.
-
- From the (draft) meeting notes:
> #### Conclusion:
> ...
I'll modify `ruby-master/tool/merger.rb tag` to use the `v4.0.0` format for Ruby 4.0+. I'll also file another ticket to announce the `v4.0.0-preview2` tag rename.
To close t... -
-
-
-
- That call is surprisingly expensive, so trying doing it once
in `#synchronize` and then passing the EC to lock and unlock
saves quite a few cycles.
Before:
```
ruby 4.0.0dev (2025-12-10T09:30:18Z master c5608ab4d7) +YJIT +PRISM [arm64-... - 10:21 PM Revision dc58d58a (git): [ruby/timeout] Fix failing timeout test
- ```
Run options:
"--ruby=./miniruby -I../ruby/lib -I. -I.ext/common ../ruby/tool/runruby.rb --extout=.ext -- --disable-gems"
--excludes-dir=../ruby/test/.excludes
--name=!/memory_leak/
--seed=9843
[ 1/31] TestTimeout#test_time... -
- I'll discuss solutions in https://github.com/ruby/timeout/issues/52.
I got an idea for a fix, but haven't had time to try it yet. -
- Fix https://github.com/Shopify/ruby/issues/874
-
- https://github.com/ruby/timeout/commit/c8d63ce3fe
-
-
- Reads better to me:
> checking prerequisites for ZJIT... yes -
-
- There are many indentation styles for AS_CASE in this file but no one
uses `[*]` for the default case. -
- Windows has no SIGUSR1.
There might be another usable signal, but this is breaking ruby master
so I just want a quick fix for now.
https://github.com/ruby/timeout/commit/b19043e8d0 -
- On the freshest `master`:
```ruby
[1].each {_1; p binding.local_variable_defined?(:_1) }
#=> true
```
It is inconsistent with changes in `#local_variables`/`#local_variable_get`/`#local_variable_set`. And inconsistent with handl... -
- With this code, we can see that the Ractor is able to access the top-level main object.
``` ruby
Ractor.alias_method :value, :take unless Ractor.method_defined?(:value)
shareable_proc = Ractor.make_shareable(Hash.new { self }.to_pro... -
- Applied in changeset commit:git|d02c97157476bbd9774f2bf6425a69166b99da1b.
----------
Stop bumping RUBY_PATCHLEVEL in release versions (#15502)
[[Misc #21770]](https://bugs.ruby-lang.org/issues/21770) -
- [[Misc #21770]](https://bugs.ruby-lang.org/issues/21770)
-
- When I filed #20959, I had #20929 in mind.
Since Windows APIs return time zone names localized based on the user's language settings rather than the system's language settings, it was important to retrieve them using a language-independ... -
- I think your idea is that Microsoft is pushing more for UTF-8 support in console apps using the Visual C++ runtime, and in the future, setting the locale to UTF-8 might become the usual way. If that happens, the console code page and loc...
-
- Just to update that though GC is still global more progress has been made. I'm on a new PC, so rerunning the benchmark against the old SHA.
```
❯ ruby -v
ruby 3.5.0dev (2025-07-02T20:01:24Z v4.0.0-preview2~2697 d5f5a56bf2) +PRISM [x... -
-
- It's likely this was fixed by the introduction of Ractor::Port. "The outgoing-port is already closed" isn't a message that can occur anymore (we no longer have outgoing ports)
-
- Help the compiler know that we always get a heap object here.
-
- This rewrites the class allocator search to be faster. Instead of using
RCLASS_SUPER, which is now even slower due to Box, we can scan the
superclasses list to find a class where the allocator is defined.
This also disallows allocating ... -
- This adds a fastpath in class_call_alloc_func to simply return if the
class matches the one expected.
I think we could probably just remove this check, or move it to the debug
build. - 04:38 PM Revision b5604833 (git): Fix Set#^ to not mutate its argument (#15296)
- * test(set): add test Set#xor does not mutate other_set
* Fix Set#^ to not mutate its argument -
-
09:45 AM Bug #21710: Segfault when reading object_id after it is set inside RUBY_INTERNAL_EVENT_NEWOBJ
- byroot (Jean Boussier) wrote in #note-24:
> > I just don't think we should support this.
> ...
To be clear, we are aware we're doing really shady weird things to get a bit less overhead / a few more features, etc and we don't expect u... -
03:35 PM Bug #21772: ruby: YJIT has panicked StackOpnd(1) should be a heap object, but was ImmSymbol for VALUE(137647867319760)
- I've attached a file the original log file was 90mb so I used
grep -B 2 -A 100 "action_dispatch/request/session.rb" /tmp/yjit_58847.log > /tmp/session_blocks.log
Hopefully that is what you are looking for. -
- It looks like you've hit the same panic as #21565.
Could you try the instructions from https://bugs.ruby-lang.org/issues/21565#note-5 to gather more information? Since you're on 3.4.7, you don't need to patch ruby for the steps there. -
- This reverts commit https://github.com/ruby/json/commit/b7e1734d9ca0.
https://github.com/ruby/json/commit/5793694ee6 -
03:01 PM Revision 4ab45e8f (git): [ruby/psych] bump snakeyaml-engine to 2.10 (jruby)
- https://github.com/ruby/psych/commit/506bf75ab2
-
-
01:38 PM Feature #21773 (Open): Support for setting encoding when a block is passed to `Net::HTTPResponse.read_body`
- Hi everyone,
This feature request could be considered a follow-up to https://bugs.ruby-lang.org/issues/2567. Opting in to automatic detection of the right encoding for the body when passing a block to `read_body` results in an error ... -
- * Fixes https://github.com/ruby/timeout/issues/17
https://github.com/ruby/timeout/commit/1a499a8f96 -
- https://github.com/ruby/timeout/commit/cb2ba88fed
-
- https://github.com/ruby/ruby/pull/15498
-
- Discussed this at the dev meeting.
I was convinced that it is very rare to use `return` or other escapes from the begin block *whose result is used* even during debugging.
```
def foo
bar(begin
return 42
1
end)
en... -
- https://github.com/ruby/timeout/commit/7cfa5a6778
-
- * Add tests related to Thread.handle_interrupt
* Fixes https://github.com/ruby/timeout/issues/41
https://github.com/ruby/timeout/commit/a52720e82a -
- For long term, believing the intelligence of the compiler sounds reasonable.
But if we believe the compiler, it doesn't need hand written parallelism. The code can be simplified like
```c
const char *
search_nonascii2(const char ... -
-
- Here is the code we found in the dev meeting:
https://github.com/ctrochalakis/mongo-ruby-driver/blob/e632eb570d2f42bddc6703c88d3ce43aea6cdfe6/lib/mongo/types/regexp_of_holding.rb#L27-L41
```ruby
class LazyRegexp < Regexp
...
... -
- We found some gems inherited from Regexp and used instance variables on instances, so they would be broken if all Regexp objects become frozen. I think it's unacceptable for Ruby 4.0.
Frankly, I don't think freezing the Regexp objects i... -
- 10:59 AM Revision 674ddf4b (git): Update default gems list at 6b469b7e40f94f25b50463de2190d5 [ci skip]
-
- https://github.com/ruby/json/commit/1cdd2122d5
-
- While it's not allowed by the spec, some parsers like Oj do
accept it, and it can be blocking a transition.
Having this feature can help people migrate.
https://github.com/ruby/json/commit/3459499cb3 -
- Applied in changeset commit:git|281a000d6669d3fbb1c0a2f940cbb9c7c01732e6.
----------
[Bug #19558] Allow ASCII range to mix with Unicode dump -
-
- Applied in changeset commit:git|dc41cf332613db24af512bc7e959fc9a17d85d59.
----------
[Misc #21690] Sync parser_bits.h from internal/bits.h -
- https://github.com/ruby/ruby/pull/15496
-
- ```ruby
re = /#{"\\p{In_Arabic}".encode("US-ASCII")}\u1234/
# encoding mismatch in dynamic regexp : US-ASCII and UTF-8
```
This behavior looks a bug. -
-
09:35 AM Bug #21712 (Closed): Prism and parse.y inconsistency in command call with block and `.()`
- Applied in changeset commit:git|c5b51bdd8c69ae038daf80cb227df854ca0110f1.
----------
[Bug #21712] Allow `.()` call for command with block
This commit allows codes like `a b do end.()` and `a b do end&.()`. -
- `parse.y` should be fixed to adopt this case.
Matz.
-
- This commit allows codes like `a b do end.()` and `a b do end&.()`.
-
- Revert "Use base: with Dir.glob for bundler.gemspec"
This reverts commit 87f0701b6106569a8486c9623dc6b0b32438355c. -
-
-
- Applied in changeset commit:git|c76ba839b153805f0498229284fea1a809308dbc.
----------
Allow String#strip etc. to take optional character selectors
[Feature #21552]
Co-Authored-By: Claude <noreply@anthropic.com> -
- I accept the proposal.
Matz.
-
- [Feature #21552]
Co-Authored-By: Claude <noreply@anthropic.com> -
- Frankly, I should have designed from the beginning that any method that accepts a block should always take a `&blk` argument, but it is impossible to change that now.
Considering that it is meaningful to explicitly state that a method... -
- I'd like to have the following new methods:
* `Binding#implicit_parameters #=> [:_1, :_2, :_3, ...] or [:it]`
* `Binding#implicit_parameter_get(:_1) #=> obj`
* `Binding#implicit_parameter_defined?(:_1) #=> true or false`
Matz.
-
- rb_process_status_new is declared in the PR as follows.
```
/**
* Creates a new instance of Process::Status.
*
* @param[in] pid The process ID.
* @param[in] status The "waitpid status", as returned by waitpid(2). This is NOT ... -
-
-
- I believe this was accidentally left in as part of
2beb3798bac52624c3170138f8ef65869f1da6c0 -
- Although the Ractor API is still experimental and may change, and there
may be some implementation issues, we should no longer say that there
are many.
Hopefully we can remove this warning entirely for Ruby 4.1
12/10/2025
-
- No point doing the manual size unit conversion for add. Sorry, no new
tests since there is no way to generate a LoadField with a negative
offset from ruby code AFAICT. Careful with the `as` casts. -
- Also in practice you'd probably want to use `sub!` to mutate in place if a big String.
That would avoid a copy, since CRuby doesn't do lazy substrings which don't share the same end. -
- Eregon (Benoit Daloze) wrote in #note-9:
> but that sounds more like something that could/should be optimized in the regexp engine
To substantiate that:
```
$ ruby -rbenchmark/ips -e 'SPACES = ["\0", *("\t".."\r"), " "].join; TARGE... -
- This sounds like a lot of complexity for one specific use-case, which already has a good solution with `sub`.
From the benchmarks, `lstrip("\0 \t-\r")` and `sub(/\A[\0\s]+/, "")` are pretty close.
`sub(/[\0\s]+\z/, "")` is slower tha... -
- tr_setup_table_multi() was called twice in String#{strip,strip!}, so I've fixed it: https://github.com/ruby/ruby/pull/15400/commits/c9cb93f201644cd5e2fbbd6e83cf50acb27642de
### Benchmark
https://gist.github.com/shugo/c6367f4139bc2d... -
- So we don't expose builders to network flakiness which cannot be worked
around using cargo's --offline flag. -
-
-
-
- 09:18 PM Revision c8909030 (git): ZJIT: Fold LoadField on frozen objects to constants (#15483)
- * ZJIT: Fold LoadField on frozen objects to constants
When accessing instance variables from frozen objects via attr_reader/
attr_accessor, fold the LoadField instruction to a constant at compile
time. This enables further optimizations... -
- Sure no problem anything else?
```
-- Ruby level backtrace information ----------------------------------------
/home/gearoid/.rvm/gems/ruby-3.4.7@7.2.3/gems/puma-7.1.0/lib/puma/thread_pool.rb:182:in 'block in spawn_thread'
/home... -
- > Backtrace & Logs
Is it possible to share the Ruby backtrace as well? -
- I'm getting a YJIT has panicked crash in my rails app locally when a part of my application rapidly uses a gem I have "auto_session_timeout".
If I patch the gem by changing the symbol to a string the error goes away.
``` ruby
... -
- Generating an object_id for any type other than T_OBJECT (and T_CLASS)
will inevitably allocate an IMEMO/fields objects, which isn't supported
in a NEWOBJ tracepoint.
See: https://bugs.ruby-lang.org/issues/21710#note-23 - 08:30 PM Revision 3640cfe5 (git): ZJIT: Use inline format args (#15482)
-
- Yeah... Either way, with the object_id now stored in `imemo/fields`, generating an object_id for anything other than `T_OBJECT` will inevitably allocate an IMEMO.
I'll remove the test case. -
- It's a noble goal, but in my view it's negative to even temporarily support this as it invites new extensions to be badly behaved and misuse these hooks.
-
- > I just don't think we should support this.
I said previously, I generally agree, but the datadog gem is quite popular, so I think we should try to find a short term solution if we can. At least that's what I tried to do. -
- This continues to be broken (in a way that I hope helps illustrate that this was always unsafe). I think there's two issues.
Since the change was introduced we're regularly seeing failures on CI related to object_id and _id2ref. Here ... -
-
05:55 PM Bug #15479: Array#reject! modifies literal Array
- Yas
-
- Encodings are RTypedData, not the deprecated RData. Although the
structures are compatible we should use the correct API. -
-
- This fixes a crash when the new shape after a transition is too complex;
we need to check that it's not complex before trying to read by index. -
-
- It's the most likely control character so it's worth
giving a better error message for it.
https://github.com/ruby/json/commit/1da3fd9233 -
- Make it embedded and compaction aware.
-
- That call is surprisingly expensive, so trying doing it once
in `#synchronize` and then passing the fiber to enter and exit
saves quite a few cycles. -
09:03 AM Bug #21771: pg-1.6.2 testsuite segfaults in ruby4.0.0dev
- Looks like this segfault is indeed fixed. Thank you!
- 07:09 AM Revision ef4490d6 (git): Update default gems list at e8a55274f202df1cfddc25aa14da34 [ci skip]
-
- https://github.com/ruby/forwardable/commit/0257b590c2
-
- Applied in changeset commit:git|14ff851185bb8ff399e98b74cc107302a4e08e18.
----------
[ruby/forwardable] Simpler and faster check for the delegation fastpath
Fix: https://github.com/ruby/forwardable/issues/35
[Bug #21708]
Trying to com... -
- Fix: https://github.com/ruby/forwardable/issues/35
[Bug #21708]
Trying to compile code to check if a method can use the delegation
fastpath is a bit wasteful and cause `RUPYOPT=-d` to be full of
misleading errors.
It's simpler and fast... -
-
- This removes all allocations from the find_or_insert loop, which
requires us to start the search over after calling the provided create
function.
In exchange that allows us to assume that all concurrent threads insert
will get the same ... -
- This refactors the concurrent set to examine and reserve a slot via CAS
with the hash, before then doing the same with the key.
This allows us to use an extra bit from the hash as a "continuation bit"
which marks whether we have ever pr... - 06:44 AM Revision 492b1c73 (git): Update default gems list at 8e87f201cf54b112642ed0421ddabd [ci skip]
-
- https://github.com/ruby/optparse/commit/f2e31e81a5
- 06:40 AM Revision a8b7fb7e (git): Update default gems list at 254653db8521618e08aaccaa63efdb [ci skip]
-
- https://github.com/ruby/win32-registry/commit/2a6ab00f67
-
- (https://github.com/ruby/stringio/pull/171)
https://github.com/ruby/stringio/commit/95a111017a -
- (https://github.com/ruby/stringio/pull/165)
Adds to "Position": pos inside a character.
Makes a couple of minor corrections.
---------
https://github.com/ruby/stringio/commit/ff332abafa
Co-authored-by: Sutou Kouhei <kou@cozmixng.org> -
- (https://github.com/ruby/stringio/pull/190)
https://github.com/ruby/stringio/commit/77209fac20 -
- (https://github.com/ruby/stringio/pull/188)
https://github.com/ruby/stringio/commit/66360ee5f1 -
- https://github.com/ruby/stringio/commit/e2d24ae8d7
-
- (https://github.com/ruby/stringio/pull/189)
https://github.com/ruby/stringio/commit/e3d16d30ed -
- Pathname is now embedded class of Ruby
- 05:30 AM Revision 74376fef (git): Update default gems list at 4523a905327d8438f845f5a7582222 [ci skip]
-
- https://github.com/ruby/date/commit/1d0aadc295
- 05:23 AM Revision ab80d05f (git): Update default gems list at 5f444cba4741b2ff0e1e95f4a17932 [ci skip]
-
- https://github.com/ruby/ipaddr/commit/93ef50bc04
-
- 05:13 AM Revision 238e69d1 (git): Update default gems list at 814f23747b5fd7b0d5fb6cd8e45833 [ci skip]
-
- https://github.com/ruby/resolv/commit/a0e89bbe48
- 05:09 AM Revision e4786376 (git): Update default gems list at df4fc0f7fcda6c552084ea0638c718 [ci skip]
-
- https://github.com/ruby/psych/commit/d8053b0d16
-
- These macros have been defined here and there, so collect them.
-
- even if `USE_RUBY_DEBUG_LOG=0`.
It becomes `fprintf(stderr, ...)`. -
- As the markers for spec/mspec/tool/remove_old_guards.rb.