Activity
From 09/05/2025 to 09/11/2025
Today
-
07:35 PM Revision f7f0f67a (git): Don't load set_spec.c if not Ruby 3.5
-
- Set was moved to C only in Ruby 3.5, so it does not compile on Ruby 3.4
and earlier. -
03:56 PM Bug #21540 (Closed): prism allows `foo && return bar` when parse.y doesn't
- Thanks to @Earlopain this is now done. https://github.com/ruby/prism/pull/3630
-
- I will make a PR for this as soon as possible to get it ready for the 3.5 release.
-
10:28 AM Bug #21540: prism allows `foo && return bar` when parse.y doesn't
- @prism This issue breaks Ruby's compatibility. This should be fixed before Ruby 3.5 (or 4.0) is released. But this is...
-
03:55 PM Revision f17405b6 (git): [ruby/prism] Reject some cases with `return` and command calls
- The same also applies to `break`/`next`.
https://bugs.ruby-lang.org/issues/21540
https://github.com/ruby/prism/comm... -
03:00 PM Revision e39fd456 (git): [ruby/prism] Use pm_arguments_end for function call
- Previously, the location of CallNode was incorrect when it accepts a
block parameter:
```
$ ruby -rprism -e 'pp Pris... -
- klass is not used, so we can shrink RZombie down to 32 bytes.
-
01:14 PM Misc #21570 (Open): Windows: Static compilation struggles
- Hi, I'm trying to make static Ruby binaries and I've discovered several issues in the **Windows** building process:
... -
12:20 PM Revision aa173bce (git): Run CAPI check separately
-
-
- rb_location_ary_to_backtrace was not checking the length of the array before
reading the first element. It can be rep... -
- The following script causes an out-of-bounds read on the $LOADED_FEATURES
array when it is modified by another thread... -
- For parameters without names, accessing the name in array index 1 would
be an out-of-bounds read. -
- We should access the last element first before we shrink the length of the
array. -
- We should assert that i is within bounds to prevent buffer overflows.
-
- If it matters only for instance variables, a new syntax won't be needed.
```ruby
instance_variable_set_unless_def... -
- Thank you for the report.
amacxz (Aleksey Maximov) wrote:
> ```
> + memcpy(&d, IBF_OBJBODY(void, offset), siz... -
08:57 AM Misc #21566: Transfer Shopify/yjit-bench and speed.yjit.org to ruby/ruby-bench and *.ruby-lang.org
- How about renaming to ruby/ruby-bench as proposed in the issue description?
Will that be done as a second step maybe... -
12:42 AM Misc #21566: Transfer Shopify/yjit-bench and speed.yjit.org to ruby/ruby-bench and *.ruby-lang.org
- @rafaelfranca transferred `Shopify/yjit-bench` to `ruby/yjit-bench` today. Thanks!
-
- Prism's behavior should be compatible with Ruby 3.3.
Unless a design change is accepted, it should not break a compa... -
03:32 AM Bug #21097: `x = a rescue b in c` and `def f = a rescue b in c` parsed differently between parse.y and prism
- I am working on this ticket and have some questions about grammar where I'd like to ask for Matz's opinions.
# Pre... -
05:21 AM Feature #21555 (Rejected): Add support for predicate attribute reader names
- I still ask you to define `def valid? = @valid` manually. It is not much worse than `attr_reader :valid?`.
Matz.
-
04:53 AM Feature #21564: Extend permutation, repeated_permutation, combination and repeated_combination arguments
- I don't think this is necessary. It's better to keep different aspects of a calculation separate by using separate me...
-
03:02 AM Feature #12282: Hash#dig! for repeated applications of Hash#fetch
- How about `dig_cons()`, as in "dig consecutive," similar to `each_cons()`? They'd be different in that the argument t...
-
- ```
1) Error:
JSONGenericObjectTest#test_from_hash:
Test::Unit::ProxyError: uninitialized constant JSON::GenericObj... -
- This reverts commit 8aa885c460aeb70878538eebdd155c6318989fd2 and
cd07c3cbae7e287350d713ead237aeef27cc2b9e. -
02:54 AM Revision e4a93944 (git): [ruby/json] Improve test coverage
- Test SubArrayWrapper#[] and SubOpenStruct#[]
https://github.com/ruby/json/commit/bae760aa19
Co-authored-by: Jean Bo... -
- https://github.com/ruby/json/commit/3232e8df41
Co-authored-by: Jean Boussier <jean.boussier@gmail.com> -
- https://github.com/ruby/json/commit/a9e05d6ff3
Co-authored-by: Jean Boussier <jean.boussier@gmail.com> -
- https://github.com/ruby/json/commit/efc61682ee
Co-authored-by: Jean Boussier <jean.boussier@gmail.com> -
- The gem root was including the test/ directory, so lines from tests were
skipped, and the deprecation warnings were s... -
- https://github.com/ruby/json/commit/7a2a24d663
Co-authored-by: Jean Boussier <jean.boussier@gmail.com> -
- Fix https://github.com/ruby/json/commit/f411ddf1ceaf
https://github.com/ruby/json/commit/1397bf8cc9
Co-authored-by:... -
- https://github.com/ruby/json/commit/f37e73cf44
Co-authored-by: Jean Boussier <jean.boussier@gmail.com> -
- (https://github.com/rubygems/rubygems/pull/8904)
* Added document for Gem::Uninstaller
* Apply suggestion from @dei... - 02:54 AM Revision 4febdbc4 (git): [rubygems/rubygems] Remove spec originally meant to test remembering CLI flags
- https://github.com/rubygems/rubygems/commit/4919e5ca53
- 02:54 AM Revision ecd1284b (git): [rubygems/rubygems] Complete remembered options removal
- https://github.com/rubygems/rubygems/commit/573ffad3ea
- 02:54 AM Revision 881b5855 (git): [rubygems/rubygems] Introduce a way to ensure that removed flags get their docs removed, too
- https://github.com/rubygems/rubygems/commit/a727b50697
- 02:54 AM Revision c218c66a (git): [rubygems/rubygems] Remove deprecated usages of `bundle install --deployment` during specs
- https://github.com/rubygems/rubygems/commit/c903afbc3f
- 02:54 AM Revision 3486f4d9 (git): [rubygems/rubygems] Fix grammar in remembered options removal message
- https://github.com/rubygems/rubygems/commit/825e29a9ec
- 02:54 AM Revision ce19fb50 (git): [rubygems/rubygems] Add missing part in `bundle cache --path` removal message
- https://github.com/rubygems/rubygems/commit/f5bf473b34
- 02:54 AM Revision 4eb13c61 (git): [rubygems/rubygems] Remove implementation of `deployment`, `capistrano` and `vlad` entrypoints
- Now they only raise an error.
https://github.com/rubygems/rubygems/commit/6e7c8db151 - 02:54 AM Revision b4334f83 (git): [rubygems/rubygems] Completely remove `Bundle.*clean*` and `Bundler.environment` helpers
- https://github.com/rubygems/rubygems/commit/e1b8bdcede
09/10/2025
-
10:35 PM Revision 9cd1ac40 (git): YJIT: Print more disassembly in release builds
- These `#[cfg(feature = "disasm")]` were unnecessary and we can provide
the information like ruby source location rega... -
-
- These `#[cfg(feature = "disasm")]` were unnecessary and we can provide
the information like ruby source location rega... -
-
09:18 PM Revision bd4de124 (git): Explicitly declare VM instruction dependencies (#14509)
- instead of mutating RubyVM::Instructions and letting the order of
`require` impact its behavior.
Now that we have bo... - 09:11 PM Revision 9401f4e7 (git): Skip `vm_cc_invalidate`'s `cc->klass` assertion for multi-ractor for now
- It's failing more frequently on CI in Ractor tests lately. To make allow
other PRs to land more smoothly, we can skip... -
- 08:37 PM Revision 23c60e18 (git): YJIT: Tiny refactors (#14505)
- Addressed some suggestions from clippy that made sense to me.
-
08:25 PM Revision 399e2abc (git): Allow concurrent_set to be collected in minor GC
- When testing we've found that the concurrent_set objects used for
fstrings can grow quite large, and because they rea... - 08:18 PM Revision 95936276 (git): ZJIT: Revert `self_val` removal and rename it to `recv` (#14504)
- * Revert "ZJIT: Removed unused self_val from Send"
This reverts commit 13c2f8d5c2b41ec78344ae60f9b5ec1564029418.
* ... -
- Thanks for trying the patch and for helping us debug! Unfortunately, we still need more information to understand the...
-
12:34 PM Bug #21565: YJIT has panicked in rails
- k0kubun (Takashi Kokubun) wrote in #note-3:
> We've modified the assertion failure message https://github.com/ruby/r... -
- 05:41 PM Revision 3e4af886 (git): ZJIT: Set ZJIT_RB_BUG to help debugging CI issues (#14500)
-
- The test introduced in 928fea3 is leaking files because the Tempfile is
not cleaned when the process exits. We see th... - 05:26 PM Revision 13c2f8d5 (git): ZJIT: Removed unused self_val from Send
- 05:26 PM Revision 877b6259 (git): ZJIT: Removed unused self_val from InvokeSuper
-
05:20 PM Bug #21569 (Open): [armv7, musl] SIGBUS in ibf_load_object_float due to unaligned VFP double load when reading IBF
- Environment:
CPU: ARMv7-A (NVIDIA Tegra 2), VFPv3-D16, no NEON
OS/libc: Linux, musl (ld-musl-armhf.so.1)
Com... - 03:57 PM Revision 33cd9700 (git): [DOC] ZJIT: Add a section for stats (#14497)
- 03:56 PM Revision d82e9f9a (git): ZJIT: Support `invokeblock` insn (#14496)
- ZJIT: Support invokeblock insn
Basically like other send-ish insns, except that it doesn't pop the receiver. -
-
-
02:49 PM Bug #21568 (Open): Requiring core libraries when already requiring multiple user defined libraries with the same name can error
- The simplest way to understand this error is that after requiring 2 or more files with the same name as some core lib...
-
- Applied in changeset commit:git|928fea3bfa86053c0bc6f7a5bf7559b115a676b5.
----------
Fix crash when $LOADED_FEATURES... -
- [Bug #21567]
When we require an object that is not a string, it will attempt to convert
it to a string by calling to... -
- Reported at https://hackerone.com/reports/3288162
This underflow does not occur in Ruby because:
* This function is... -
-
-
- 05:38 AM Revision 6a4222f4 (git): Update default gems list at 6c9a0173332121f15fa0c654b1a728 [ci skip]
-
- https://github.com/ruby/fcntl/commit/13d8aa4785
-
04:42 AM Revision 37884c4a (git): [ruby/fcntl] Add macOS specific flags around file preallocation
- I wanted to use file preallocation with fcntl, but the flags weren't
available. This commit just adds the missing fl...
09/09/2025
-
-
- This reverts commit 9d0b621b051c0f3cc356c9bedf081456f1d7dfd6.
It didn't work for `make check`. I'll fix it in a PR. -
- ref: https://github.com/ruby/ruby/pull/14459
- 08:25 PM Revision d14b4a70 (git): ZJIT: Remove unneeded omission for optional param test (#14488)
-
-
-
-
- GitHub PR: https://github.com/ruby/ruby/pull/14487
When we require an object that is not a string, it will attempt... -
05:33 PM Revision 9e3a1622 (git): [DOC] Tweaks for String#lstrip!
-
- Co-authored-by: Alan Wu <XrXr@users.noreply.github.com>
-
03:58 PM Revision f7fe4361 (git): ZJIT: Optimize `ObjToString` with type guards (#14469)
- * failing test for ObjToString optimization with GuardType
* profile ObjToString receiver and rewrite with guard
* ... -
03:55 PM Revision fe362beb (git): ZJIT: Avoid mutating string in zjit stats (#14485)
- [ZJIT] Avoid mutating string in zjit stats
GitHub runs with a Symbol patch that causes a frozen string error when
ru... -
- If we alias two variable multiple times, for example like this:
alias $foo $bar
alias $foo $bar
Then we wil... -
-
07:15 AM Bug #21560: RUBY_MN_THREADS=1 causes large performance regression in Puma 7
- Seems not limited to `sleep`, all scheduling seem impacted:
```ruby
th = Thread.new { sleep 10 }
Thread.new do... - 04:53 AM Revision 5c98b899 (git): Bump actions/github-script from 7.0.1 to 8.0.0
- Bumps [actions/github-script](https://github.com/actions/github-script) from 7.0.1 to 8.0.0.
- [Release notes](https:... - 04:01 AM Revision 89bce815 (git): Bump actions/checkout from 4 to 5
- Bumps [actions/checkout](https://github.com/actions/checkout) from 4 to 5.
- [Release notes](https://github.com/actio... - 03:22 AM Revision 27d9c67b (git): Bump actions/checkout in /.github/actions/setup/directories
- Bumps [actions/checkout](https://github.com/actions/checkout) from 4.2.2 to 5.0.0.
- [Release notes](https://github.c... -
-
- Sounds great! Go ahead.
Matz.
09/08/2025
-
- ## Proposal
1. Transfer `Shopify/yjit-bench` to `ruby/ruby-bench` on GitHub
* Let `@ruby/jit` (YJIT/ZJIT deve... -
- entry1 is allocated using xmalloc (through ALLOC), which returns undefined
memory. We never set the ractor_local fiel... -
10:27 PM Feature #21564: Extend permutation, repeated_permutation, combination and repeated_combination arguments
- @nobu I was not sure what to select, Enumerable or multiple arguments, changed the proposal to use multiple arguments...
-
- When proposing new features, I strongly recommend writing a use case. It should be as simple and relatable as possibl...
-
- I'm not sure whether this is a natural extension of permutation.
And why not `array.permutation(count1, count2, co... -
- We've modified the assertion failure message https://github.com/ruby/ruby/pull/14480 to debug this further. I'll rele...
-
-
- I've got YJIT has panicked crash in rails app locally on my M1 Pro Mac Book pro for 2 times for last 2 weeks
```
... -
- YJIT: Add more information to an assert message in jit_guard_known_class (#14480)
---------
Co-authored-by: Alan Wu... -
-
- ZJIT, YJIT: Make the workflow names consistent
with file names -
- I suspect this might be the issue: it seems like under M:N threads sleeps < 10ms are rounded up.
``` ruby
Thread.... -
07:11 PM Revision 4263b7eb (git): ZJIT: Add RubyVM::ZJIT.reset_stats! method (GH-14479)
- This allows for more precise tracking of stats programmatically
which is particularly useful for our nightly benchmar... -
06:56 PM Revision 80079ba4 (git): ZJIT: Fix 30k if stmt test (#14446)
- * ZJIT: Allow label generation above 19 bits
* Refactor emit_conditional_jump to use generate_branch
* Make branchin... -
- * YJIT: Add more information to an assert message
in jit_guard_known_class
* Say "should be a heap object" instead
... -
- @ttilberg That's right. The backtrace has `bind_call` as the caller of `binding`, and `bind_call` is a C method, henc...
-
05:56 PM Bug #18487: Kernel#binding behaves differently depending on implementation language of items on the stack
- I'm upgrading a project from Ruby 3.0 to Ruby 3.4, and I think I came across an unexpected bug related to this relate...
-
-
- Followup changes in https://github.com/ruby/ruby/pull/14470 /
03c86b053197f3cd6bece1925e634c1d74d196d0 -
- Similar to YJIT's --yjit-stats=quiet, this option allows ZJIT to collect
statistics and make them available via the R... -
- Followup: https://github.com/ruby/ruby/pull/14463
-
- Probably due to the testing order, sometimes it looks like that
`Gem::UnknownCommandError` happens to be used without... -
- * Revert "Debug: More inspections"
e5a6e952467eef5444eaf63b77d8c91aa367d4de.
* Revert "Debug for sporadical failur... -
- > The source problem is the encoding US-ASCII for "."
Yes, but `File.expand_path` behavior is also suspicious.
... -
- Ref: https://github.com/ruby/ruby/pull/14134#issuecomment-3207733725
We can't safely use low-bit pointer tagging any... -
11:23 AM Bug #19288: Ractor JSON parsing significantly slower than linear parsing
- I can confirm significant performance improvements with the new Ractor APIs. I will now proceed to benchmark this aga...
-
-
09/07/2025
-
- When using functions `permutation`, `repeated_permutation`, `combination` and `repeated_combination`, often one needs...
- 03:41 PM Revision 1e7ee6a4 (git): [DOC] Improve format specification docs
- One example to describe how `*` works actually prints a warning:
```
$ ruby -we "sprintf('%d', 20, 14)"
=> -e:1: war... -
03:23 PM Revision 6cd98c24 (git): Allow IO#each_codepoint to work with unetc even when encoding conversion active
- Using IO#each_codepoint together with IO#ungetc causes an unwanted exception when encoding conversion is active.
C:\... -
09/06/2025
-
03:52 PM Bug #21563 (Closed): Misleading error message when `to_proc` does not return a Proc in an Object used as a &block argument
- Applied in changeset commit:git|953e1ef99283d8563ff655ee6b8fcd681af79c1c.
----------
Make invalid & operator type er... -
06:58 AM Bug #21563: Misleading error message when `to_proc` does not return a Proc in an Object used as a &block argument
- I submitted a PR to make the error messages for & consistent with */**: https://github.com/ruby/ruby/pull/14463
-
- If we want to change the behavior in this case, we should report both the actual argument and the return value of `to...
-
- If #to_proc is defined, this uses the following error message format,
matching the error message format used for * wh... -
- - Remove unnecessary string interpolations.
- `/#{SEPARATOR_PAT}/o` is always same as `SEPARATOR_PAT` Regexp. -
-
-
- PR here
https://github.com/ruby/ruby/pull/14455
09/05/2025
-
11:56 PM Bug #21563 (Closed): Misleading error message when `to_proc` does not return a Proc in an Object used as a &block argument
- When a class implements `#to_proc` it should always return a Proc. Nonetheless bugs are possible, and when an impleme...
-
-
-
-
- The previous one failed:
https://github.com/ruby/ruby/actions/runs/17504307509/job/49724437021
$GITHUB_ACTION_REPOSI... -
-
- This fails way too often these days, probably since
https://github.com/rubygems/rubygems/pull/8948.
To make sure we ... -
- 08:08 PM Revision c06e7046 (git): CI: ubuntu-ibm.yml: Set the runner user's primary group from the "id -g".
- Set the runner user's primary group to avoid a mismatch between the group IDs of
"id -g" and C function getpwuid(uid_... - 08:08 PM Revision 76deabd3 (git): CI: ubuntu-ibm.yml: Fix a typo to print HOME env.
-
- if it's the only JIT enabled in the build
-
-
- For example, `defined?(yield)` never calls a method, so it's leaf.
-
- 83bc6ced920c51ffc3783b03f2e1c7f89f5c2fd5 was the last use of it.
-
- It was used to let MJIT override the leafness of the instruction when it
decides to remove check_ints for it. Now tha... - 04:07 PM Revision 6c88b719 (git): ZJIT: Remove unneeded tracing and thread test exclusions (#14453)
-
- Found by wbcheck
-
- Found by wbcheck
-
- Found by wbcheck
-
- Found by wbcheck
-
- Found by wbcheck
-
- Found by wbcheck
-
- Found by wbcheck
-
- We want to always use time_set_timew, as timew is 64-bit even on 32-bit
platforms so we need to be careful to both wr... -
-
-
- We are inserting directly into the st_table, so we need to issue a write
barrier from the hash. -
- This write barrier occurred before the entry was added to the table, so
if GC occurred when inserting into the table,... -
- After returning from the callback in st_update is the point that the
hash table may be resized, which could trigger a... - 09:32 AM Revision 0d91997c (git): [rubygems/rubygems] Improve error message when `path` and `gemspec` point to the same gem in Gemfile
- https://github.com/rubygems/rubygems/commit/e47a9064be
- 09:32 AM Revision a9db92f9 (git): [rubygems/rubygems] Add gem sources --prepend and --append for finer source control
- --append adds a source to the end, moving it to the end if it already exists.
--prepend adds or moves a source to the... -
09:32 AM Revision 41ffd1bd (git): [rubygems/rubygems] Fix "did you mean" suggestions for unknown commands
- Since Ruby 3.4.5, which ships with did_you_mean-2.0.0, RubyGems no
longer gives "did you mean" suggestions for unknow... -
- This reverts commit 334d49376dd1afca7427d5e2070d5cc9e270361a.
Seems just `LC_CTYPE` was set to a wrong value. -
- Workaround for a very strange behavior in macOS (reported about 10.15
at least): `nl_langinfo(CODESET)` returns an em...