Activity
From 04/29/2026 to 05/05/2026
Today
-
01:09 AM Revision cc31e846 (git): Add write barriers to ARGF
-
12:10 AM Revision 06fc5c24 (git): Use EC saved in GC for root marking
- Since EC is thread-local, we previously used rb_gc_worker_thread_set_vm_context
in MMTk worker threads to temporarily set the EC. However, this was inelegant
and also occasionally caused crashes when marking threads/fibers for the
curren...
05/04/2026
-
10:06 PM Revision 1ce2e79b (git): - iseq.c: fix passing frozen option to compile_file_prism
-
10:06 PM Revision 43c01a5f (git): ISeq.compile_file was not switching on parser
-
10:06 PM Revision 4c564ee2 (git): Fix coverage support for RubyVM::ISeq.compile
- [Bug #22018]
ISeq returned by `RubyVM::InstructionSequene.load_iseq` weren't
handled by the coverage module. -
09:45 PM Revision 138b4ba2 (git): ZJIT: Skip heap-upgrade snapshot on unsupported builds
- Stop deriving the embedded-capacity boundary from GC internals in
`upgrade_self_type_to_heap_after_setivar`.
Instead, use self-consistency checks to keep the snapshot only on builds
where five ivars stay embedded and the next write over... -
- The old test snapshot depended on the second write crossing the
embedded-capacity boundary, even though the behavior under test was
just the shape transition across two ivar writes.
Warm the class to two ivars and assert the natural spe... -
- The old setivar opt tests depended on embedded-capacity boundaries that
vary between dev and stats builds.
Compute the boundary from GC::INTERNAL_CONSTANTS and construct receivers
so the tests reliably cover shape transition and heap-ba... -
- Avoid receiver shapes near the embedded/heap boundary in the
polymorphic getivar tests.
Make both receivers clearly heap-backed so stats and dev builds still
exercise distinct @foo slots without depending on embedded_p(). -
09:08 PM Bug #22020: Inner call node without all arguments returned by RubyVM::AbstractSyntaxTree.of for call with a block
- My motivation here is I would like to implement `Thread::Bactrace::Location#source_range`.
For this to work in `--parser=parse.y` mode it needs to return the same start/end line/column as Prism in all cases of `RubyVM::AbstractSyntaxTre... -
- The current locations are like this:
```
foo(1, 2, kw: :arg) { 123 }
|---------ITER------------|
|------FCALL------| |SCOPE|
```
When `foo` does not exist, it's a `NoMethodError`, and `RubyVM::AbstractSyntaxTree.of e.backtrace_locations... -
- I think this is safe to reject, I think it's not feasible to change this.
One should avoid mutating the Array while iterating, otherwise this kind of behavior is expected. -
- IMO it's not reasonable for `Array#delete_if` to do anything else than what it already does.
How can `Array#delete_if` know which element you deleted and whether it should delete or not?
The block returns true for the second 2nd block ca... -
05:52 AM Bug #22021 (Rejected): Array#delete_if may delete wrong object if array has been altered already
- The simplest example I can come up with:
```ruby
$ar = ar = [1, 2, 3, 4, 5]
def del(i)
$ar.delete(i)
end
ar.delete_if { |e| e == 2 ? (del(e) && true) : false }
p ar #=> [1, 4, 5], and it should be [1, 3, 4, 5]
``` -
- Backport PRs:
- 4.0: https://github.com/ruby/ruby/pull/16852
- 3.4: https://github.com/ruby/ruby/pull/16853
- -
- - Fix passing frozen_string_literal option to #compile_file backed by Prism: https://github.com/ruby/ruby/pull/16779
- ISeq.compile_file was not switching on parser: https://github.com/ruby/ruby/pull/16463
-
-
- They're the same code.
-
- By only storing shape offsets in the cache, we're able to still
match two equal objects that happen to be in different heaps, as
well as to validate the object is neither frozen nor complex. -
- It's the same code duplicated in 3 places.
-
03:22 PM Revision 1aa62e89 (git): [ruby/prism] Optimize ripper `visit_token`
- It was showing up in profiles.
So:
* Don't splat `KEYWORDS` (also did the same for `BINARY_OPERATORS`)
* Use `start_with?` if possible
Overall gives a ~5% speed boost
https://github.com/ruby/prism/commit/d611aa9d11 -
- It's a small, somewhat hacky performance boost. Locations are lazy, by freezing the
result they don't have to be pack/unpacked redundantly.
This gives about a 4% speed boost.
Other changes are to not modify the frozen AST
https://githu... -
- Basically a port of https://github.com/ruby/ruby/commit/c45f781771314a71856c9b348c640ba532f54349 into ruby
It's quite effective at ~97% hit rate for me.
Speeds it up from ~6.77x slower to only 4.07x slower.
For the lexer `on_sp` it als... -
-
-
12:33 PM Bug #22004: parse.y doesn't executes loop body with `while true || true` condition
- Backported into ruby_3_4 at commit:fdac94eee1a3f226cafcfa6ec0da97bb08b03369
-
- [PATCH] [Bug #22004] Fix short-circuited loop conditions
-
- Backports:
- 4.0: https://github.com/ruby/ruby/pull/16846
- 3.4: https://github.com/ruby/ruby/pull/16847 -
- Applied in changeset commit:git|8bf6d7bfcb4ca46d2cfe73bdbce2fdb0652b1aae.
----------
Fix coverage support for RubyVM::ISeq.compile
[Bug #22018]
ISeq returned by `RubyVM::InstructionSequene.load_iseq` weren't
handled by the coverage mo... -
- [Bug #22018]
ISeq returned by `RubyVM::InstructionSequene.load_iseq` weren't
handled by the coverage module. -
- I understand the reasoning, however given the implementation of boxes, I fear implementing this proposal would reintroduce the need for many ractor locks, negating their usefulness.
Also it's not just `Ruby::Box` that would need to be... -
- Since more size pools have been added, this constant was outdated.
-
- We ran into several more instances of that bug in the last couple weeks.
```
apps/intercom/app/helpers/operator_workflows_helper.rb: [BUG] Segmentation fault at 0x0000aaaa00000009
ruby 4.0.2 (2026-03-17 revision d3da9fec82) +PRISM [... -
06:29 AM Bug #22022 (Open): Refinement zsuper problems when referenced method changes
- I've found many problems with refinement zsuper methods. Here's one example:
```ruby
module M
private def a = :a
alias a a
end
class A
include M
end
class B < A
end
module R
refine B do
public :a
end
... - 04:58 AM Revision ca4047c5 (git): Update default gems list at 5ce5cc8ee24ad92890507a1b0d0c68 [ci skip]
-
- https://github.com/ruby/json/commit/4a1a4a4861
-
- Caught by wbcheck. The same field is written with RB_OBJ_WRITE
in rb_method_definition_set. -
-
- Initial implementation of the wbcheck GC, a modular GC that detects
missed write barriers by collecting references from each object and
comparing against the set of objects that were marked/written via
rb_gc_writebarrier. -
02:50 AM Revision 5d439f29 (git): mswin: Suppress quirky warning
- Passing an array of const pointers within a struct as an argument
seems to be too complex for cl.exe:
```
vm_eval.c(501): warning C4090: 'function': different 'const' qualifiers
``` - 02:31 AM Revision c859018a (git): Bump the github-actions group across 1 directory with 2 updates
- Bumps the github-actions group with 2 updates in the / directory: [github/codeql-action](https://github.com/github/codeql-action) and [taiki-e/install-action](https://github.com/taiki-e/install-action).
Updates `github/codeql-action` f... -
-
05/03/2026
-
-
11:15 PM Revision 3d7a5678 (git): [DOC] Doc for Pathname#read
-
- (https://github.com/ruby/strscan/pull/203)
They don't work in included markdown and also need a leading `:`
<img width="791" height="442" alt="grafik"
src="https://github.com/user-attachments/assets/eca5d4c2-9597-4c58-a0cc-a4e2c9109992... -
- It doesn't impact ivar lookup, because OBJ_ID fields change
the offset. -
- It better describe that part of a `shape_id`.
-
- It's probably the most helpful field while debugging, having
it toward the start of the struct ensure `rp` and other debugger
helpers will display it. -
- When the shape doesn't have a SHAPE_OBJ_ID edge, we don't
really have to rebuild it, we can simply combine the old shape
offset with the new shape flags. -
- ```ruby
begin
foo(1, 2, kw: :arg) { 42 }
rescue => e
pp RubyVM::AbstractSyntaxTree.of e.backtrace_locations[0]
end
```
```
$ ruby --parser=parse.y rubyvm_ast_node_id_loc.rb
(FCALL@2:2-2:21 :foo
(LIST@2:6-2:20 (INTEGER@... -
- Avoids the need to cast fields into a st_table.
-
- Applied in changeset commit:git|7f4f3c8ee7cde795dc27be753796fc7ea8318565.
----------
[Bug #22004] Fix short-circuited loop conditions -
-
- @tikkss I think there is an issue with your accounting of memory because it doesn't match the RSS of the worker process.
This is running your example as-is.
```
$ ruby -v
ruby 4.0.2 (2026-03-17 revision d3da9fec82) +PRISM [x86_64-linux]... -
- Expose both `rb_obj_shape_` functions that take a `VALUE`
and `rb_shape_` functions that take a `shape_id`.
Make common transition functions such as `complex` and `frozen`
inlineable.
Also get rid of RB_SET_SHAPE_ID and rb_set_boxed_cl... -
- We don't need to call into `shape_transition_object_id`, we know
we're transitioning to TOO_COMPLEX so the offset part of the id
can only be `ROOT_TOO_COMPLEX_SHAPE_ID` or `ROOT_TOO_COMPLEX_WITH_OBJ_ID`.
So there's no need to lookup the... -
05:56 AM Bug #21870: Regexp: Warnings when using slightly overlapping \p{...} classes
- Hi there! It's been a few weeks and I want to make sure this doesn't get lost - please understand that despite how it looks from a cursory viewing of the diff, the patch I've submitted does seem to remove the warning from exactly this ca...
05/02/2026
-
- Extracted from: https://github.com/ruby/ruby/pull/16817
It's likely that it will be u8 soon. -
03:02 PM Feature #21953: Allow accessing unshareable objects within a Ractor-local Ruby Box
- Basically, I agree to this proposal. The described use case is very clear and understandable to me.
Technically, we can find several things to be considered.
1) It's not clear how a box is tied with a ractor
This proposal is bas... -
- Yes
-
12:43 PM Feature #21998: Add {Method,UnboundMethod,Proc}#source_range
- Eregon (Benoit Daloze) wrote in #note-21:
> I confirmed with both @mame and @matz, and matz said it should be from `p` of `proc`.
So that means it will include the original call and all arguments in all cases, including multi-line ca... -
- Eregon (Benoit Daloze) wrote in #note-10:
> I'm not sure if [@matz (Yukihiro Matsumoto)](/users/13) made his mind on: Start position: from the `p` of `proc`? Or from the `{`?
I confirmed with both @mame and @matz, and matz said it sh... -
- This shrinks the `rb_shape_t` struct from 40B to 32B.
-
-
- Embed and shrink the capacities array so that it can be queried
more efficiently.
Move `next_shape_id` and `cache_size` outside of the struct on their
own cache line, as they're expected to be incremented atomically from
concurrent thre... -
- Simplify the API by taking a shape_id upon creation.
-
- Otherwise some code defining ivars on a naked object early during
program boot can cause all objects to be larger than needed.
This seem like it was always the intention, but wasn't quite
properly prevented.
Also stop updating `max_iv_... -
05:02 AM Bug #22015: Success without RUBY_BOX=1, Failure with RUBY_BOX=1
- I found a minimal reproduction case.
The issue seems to relate to how `Enumerable#flat_map` handles `Symbol#to_proc` within a `Ruby::Box` environment. Notably, `map(&:method)` works as expected, while `flat_map(&:method)` fails.
```...
05/01/2026
-
- It was dropped in https://github.com/ruby/ruby/commit/932dd9f10e684fa99b059054fbc934607d85b45a
-
- This adds specialization for Class#allocate to perform the same inlining
that Class#new will do. This was done conservatively, only inlining when
we see a known leaf allocator so that there's no difference between
interpreter and JIT if ... -
- ```ruby
class C
include Enumerable
def each(&b)
$b = b
yield 1
end
end
Set[1, 2, 3] & C.new
$b.call(1) # [BUG] Segmentation fault at 0x00007f21bfa67f60
```
The cause is essentially the same as #5801, we're in... -
12:01 PM Feature #21953: Allow accessing unshareable objects within a Ractor-local Ruby Box
- Eregon (Benoit Daloze) wrote in #note-2:
> Did you measure the difference in memory usage?
> ...
Thanks for your reply. No, I didn't. But I took this opportunity to give it a try.
I increased the number of classes with three methods to ... -
- https://github.com/ruby/json/commit/f6ca597cb1
-
03:10 AM Bug #21940 (Closed): Ruby::Box: `$_` returns stale value due to gvar_tbl caching
- Applied in changeset commit:git|f89b07ef0046257dd796a2e615cc063072114f16.
----------
Mark `$_` as box-dynamic to bypass Box gvar_tbl cache
`$_` is updated through svar (rb_lastline_set), bypassing
rb_gvar_set, so the Box gvar_tbl cache... -
- `$_` is updated through svar (rb_lastline_set), bypassing
rb_gvar_set, so the Box gvar_tbl cache is never invalidated
and returns a stale value. Call rb_gvar_box_dynamic so
gvar_use_box_tbl() skips the cache.
Fixes [Bug #21940](https://... - 02:41 AM Revision 81f2fcae (git): Bump insta in /zjit in the jit group across 1 directory
- Bumps the jit group with 1 update in the /zjit directory: [insta](https://github.com/mitsuhiko/insta).
Updates `insta` from 1.47.1 to 1.47.2
- [Release notes](https://github.com/mitsuhiko/insta/releases)
- [Changelog](https://github.co... -
01:50 AM Revision 5219c4c8 (git): [ruby/rubygems] Bump rb_sys to >= 0.9.127
- https://github.com/ruby/rubygems/commit/cf21e9113f
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
04/30/2026
-
-
-
- * [Feature #21981] Remove CREF rewriting for methods on cloned classes/modules
* Current behaviour is inconsistent with all other constant references being lexically scoped (inheritance, mixins, define_method). Only `Class#{dup,clone}... -
- onig_foreach_name is totally synchronous (either a for loop or an st_foreach), so it should be safe to pass data on the stack rather than allocating an IMEMO.
-
-
- We always know when this contains a VALUE and even fire a write barrier
for it. We should use this when marking, support compaction, and no
longer need conservative marking. -
- This was no longer used anywhere
- 02:27 PM Revision df623125 (git): [ruby/prism] Fix UTF-8 code units to match the number of bytes
- (https://github.com/ruby/prism/pull/4098)
https://github.com/ruby/prism/commit/442bd904ed -
-
- It's here https://github.com/ruby/ruby/pull/16434
4.0.3 was a security release only, so you can expect to get this fix in 4.0.4 which is intended to be released in may -
04:33 AM Bug #21856: Massive performance degradation of `rb_obj_free` for `T_CLASS` since Ruby 4.0
- Could we possibly check on this again? I checked master and found the change, but couldn't find it in neither 4.0.2 nor in 4.0.3. I looked in the repository on github as well into the official 4.0.3 tarball published on ruby-lang.org. It...
-
-
- The blanket skip in checksums_section from #9506 caused 82 ruby-core test
failures whose lockfile expectations legitimately include the bundler
checksum (paths that go through `bundle install` where bundler is
activated as a gem). Revert... - 03:12 AM Revision 5dcc311f (git): enumerator.c: guard eargs during inspect
-
- fixtures
(https://github.com/ruby/rubygems/pull/9506)
* Skip bundler self-checksum on ruby-core in test fixtures
In ruby-core test setup, bundler is loaded as a default gem and the
bundler.gem cache file is not present on disk where
Lo... - 02:24 AM Revision 7dda33be (git): Bump taiki-e/install-action
- Bumps the github-actions group with 1 update in the / directory: [taiki-e/install-action](https://github.com/taiki-e/install-action).
Updates `taiki-e/install-action` from 2.75.25 to 2.75.27
- [Release notes](https://github.com/taiki-e... -
- (https://github.com/ruby/rubygems/pull/9501)
* Skip bundler self-checksum for unreleased bundlers
Using `Bundler.gem_version.end_with?(".dev")` only skips the own
checksum on master, but patch releases run from a source checkout
(e.g.,... -
12:08 AM Revision d61424bb (git): ZJIT: More recognizable JIT return poison, assert only when RUBY_DEBUG
- A good poison value gives a recognizable fault address if someone
erroneously dereferences. Asserting the pointer is not poison shouldn't
be necessary, so move it to RUBY_DEBUG builds.
04/29/2026
- 10:24 PM Revision 9f24ba51 (git): ZJIT: Specialize direct sends to methods with post-required positional parameters
-
- When references show up on in the type declaration, it's an invitation
to think about how long the implicit lifetime is. This code
doesn't do anything tricky lifetime-wise, so it looks better declaring
only raw pointers. -
- As the documentation puts it, transmute is "incredibly unsafe". We can
express what we need to with pointer casts in these closure FFI
situations, so let's use pointer casts. -
- As the documentation puts it, transmute is "incredibly unsafe". We can
express what we need to with pointer casts in these closure FFI
situations, so let's use pointer casts. -
- As well as `putchilledstring` as `dupchilledstring`.
This is more consistent with similar `duparray` and `duphash`
instructions and better reflect it's behavior. -
- Oh! Looks like this was already fixed by https://github.com/ruby/ruby/pull/16792
-
- That makes sense that this crashes. We're allowing the GC to provide xmalloc and xfree, and in Ruby 4.1dev under debug that has additional padding. We should have an xmalloc equivalent of malloc_usable_size (at least in the one place we ...
-
- For
```ruby
l = -> x, y {
x + y
}
l.call(1)
```
ErrorHighlight shows:
```
test.rb:1:in 'block in <main>': wrong number of arguments (given 1, expected 2) (ArgumentError)
caller: test.rb:5
| l.call(1)
^^^^... -
- I was curious what's the reported location for `ArgumentError`s caused by lambda parameters and how ErrorHighlight highlights them.
For
```ruby
l = lambda { |x, y|
x + y
}
l.call(1)
```
With `parse.y`, the node (`RubyVM::... -
05:58 PM Bug #22016: ruby segfaults regularly
- Oh, sorry, you asked for more information:
uname -a was in the original bug report, I'm using NetBSD 11.99.5 on x86_64 (= amd64).
-
- Ruby does not compile out of the box for me, I'll attach the build log. I've attached the build log of a plain ruby build outside of pkgsrc, same system.
I don't know why it uses clang by default, 'cc' is gcc 14.3.0.
In case you're i... -
05:16 PM Bug #22016: ruby segfaults regularly
- Can you please provide the output of `lscpu` and/or `uname -a` along with the version of `gcc`? What happens when you install this exact Ruby version without pkgsrc, like from git or a tarball? Thanks.
-
01:13 PM Revision 3f75e999 (git): Bump RDoc to latest master (bc0baee) (#16807)
-
12:25 PM Revision f1b2b6ba (git): [ruby/openssl] ssl: refactor SSLSocket#{peer_,}finished_message
- The variable-sized alloca makes me nervous, even though it turned out
to be safe: the Finished message is 36 bytes long in SSL 3.0 and is
usually smaller in newer protocol versions. However, the alloca is not
actually needed since we can... - 11:15 AM Revision 686b071a (git): [ruby/openssl] pkcs7: fix error check of PKCS7_SIGNER_INFO_set()
- This function returns <=0 on error, not only 0.
This is evident by looking at the implementation where it can return a
negative error here [1]. Use in OpenSSL itself also uses the <=0 check [2].
[1] https://github.com/openssl/openssl/bl... - 11:08 AM Revision a130f137 (git): [ruby/openssl] asn1: add error checks for ASN1_STRING_set() calls
- Other calls in the project have such checks, add it consistently
everywhere to raise an exception in case of an error.
https://github.com/ruby/openssl/commit/ac48b6c8af -
- Fix CI with OpenSSL master in the FIPS mode. With the fips provider in
OpenSSL master, HMAC keys shorter than 112 bits are rejected and MD5 is
disallowed. Update tests to use SHA-2 and longer keys.
https://github.com/ruby/openssl/commit... -
- Fix-up for d3ef85a100a4d051fcafc6ef6c09c4faaecc086a.
- 09:59 AM Revision e3fe84d2 (git): [ruby/openssl] x509attr: use sk_ASN1_TYPE_new_reserve()
- This should avoid reallocations and prevent the main error condition of
the push call.
https://github.com/ruby/openssl/commit/0966d0466e - 09:59 AM Revision 801fc083 (git): [ruby/openssl] x509attr: check for errors of sk_ASN1_TYPE_push()
- This function returns 0 on error.
https://github.com/ruby/openssl/commit/8ca0afb30a - 09:56 AM Revision a9cbdceb (git): [ruby/openssl] cipher: fix error checks for EVP_CIPHER_CTX_ctrl()
- The return value behaviour isn't explicitly documented for OpenSSL.
For LibreSSL, it says [1]:
> EVP_CIPHER_CTX_ctrl() returns 1 for success or 0 for failure.
> ...
So it appears that we need to check for `<= 0` instead of `!0`.
Further... -
- Reproduction:
```ruby
require "coverage"
File.write("/tmp/a.rb", <<~RUBY)
module CoverableRaw
def self.call
"cover up"
end
end
CoverableRaw.call
RUBY
Coverage.start
require "/tmp/a.rb"
p Coverage.... -
- * In file included from /usr/include/string.h:535,
from include/prism/internal/arena.h:12,
from src/prism.c:6:
In function 'memset',
inlined from 'lex_mode_push_regexp' at src/prism.c:290:5:
.../... - 07:48 AM Revision eadefdd0 (git): Introduce `IO::Buffer#bit_count`. (#16784)
-
- As of Ruby 4.0, the YJIT comment isn't quite correct because we now
store the reference to the `imemo/fields` object inline.
Which means that a Struct of precisely 78 members (max embeddable on
64bit archs) can cohexist in embedded and ... - 02:25 AM Revision 23215de7 (git): Bump taiki-e/install-action
- Bumps the github-actions group with 1 update in the / directory: [taiki-e/install-action](https://github.com/taiki-e/install-action).
Updates `taiki-e/install-action` from 2.75.23 to 2.75.25
- [Release notes](https://github.com/taiki-e... -
02:09 AM Revision 493d26dc (git): Reject overflowed repeat/null-check IDs in huge regexps (#16801)
- Add explicit errors for excessive range-repeat and null-check IDs;
map new errors in regerror; add regression tests for oversized generated
patterns. -
- > emulate this feature by using autoload :Constant, "file/that/raises/exception".
We also suggested studying the possibility of allowing "autoload blocks":
```ruby
autoload :Constant { raise "Constant shouldn't be accessed from #{...