Activity
From 06/22/2024 to 06/28/2024
06/28/2024
-
08:04 PM Revision 291cc913 (git): [Bug #20598] Fix corruption of internal encoding string (#11069)
- Fix corruption of internal encoding string
[Bug #20598]
Just like [Bug #20595], Encoding#name_list and Encoding#aliases can have
their strings corrupted when Encoding.default_internal is set to nil.
Co-authored-by: Matthew Valentine-H... -
05:44 PM Revision 3e14fe7c (git): YJIT: Fix `cargo doc --document-private-items` warnings [ci skip]
- Mostly putting angle brackets around links to follow markdown syntax.
-
- Many functions take an outlined code block but do nothing more than
passing it along; only a couple of functions actually make use of it.
So, in most cases the `ocb` parameter is just boilerplate.
Most functions that take `ocb` already ... -
02:38 PM Revision 01f0dcd3 (git): Fix missing macOS version in Slack notifications
- matrix.os is not set for some jobs.
-
02:12 PM Revision 98c923ff (git): Synchronize Bundler & RubyGems (#11071)
- 02:11 PM Revision 1652c194 (git): Fix comment for VM_CALL_ARGS_SIMPLE (#11067)
- * Set VM_CALL_KWARG flag first and reuse it to avoid checking kw_arg twice
* Fix comment for VM_CALL_ARGS_SIMPLE
* Make VM_CALL_ARGS_SIMPLE set-site match its comment -
- https://github.com/rubygems/rubygems/commit/88e3f1d23c
-
- https://github.com/rubygems/rubygems/commit/91274128a8
-
- https://github.com/rubygems/rubygems/commit/8f34396af6
-
- Instead, build it during setup when in CI.
This should avoid some Windows specific test failures when Bundler
copies the same files from multiple processes and runs into EACESS
errors.
https://github.com/rubygems/rubygems/commit/c194a1... -
06:52 AM Bug #20588 (Closed): RangeError: integer 132186463059104 too big to convert to 'int' since cdf33ed5f37f9649c482c3ba1d245f0d80ac01ce with YJIT enabled
06/27/2024
-
11:58 PM Bug #20599 (Closed): TypeError: #<BaseMailer:0x00000000008070> is not a symbol nor a string since 4cbc41d5e5cb6793174d5964975fdb4470323ca1 with YJIT enabled
https://buildkite.com/rails/rails-nightly/builds/711#01905b80-6b82-460c-908d-c8bc7e69293a/1213-1221
### Steps to reproduce
```
git clone https://github.com/rails/rails
cd rails/actionmailer
bundle install
RUBY_YJIT_ENABLE=1 bin...-
- It looks like 4cbc41d5e5cb6793174d5964975fdb4470323ca1 addresses this issue.
```
$ RUBY_YJIT_ENABLE=1 bin/test test/cases/scoping/relation_scoping_test.rb test/cases/associations/eager_test.rb --seed 1670
/home/yahonda/.gem/ruby/3.4... -
- Applied in changeset commit:git|176c4bb3c7db87ca5b0486012cb6a005105448c5.
----------
Fix corruption of internal encoding string
[Bug #20598]
Just like [Bug #20595], Encoding#name_list and Encoding#aliases can have
their strings corrup... -
- GitHub PR: https://github.com/ruby/ruby/pull/11068
Just like #20595, Encoding#name_list and Encoding#aliases can have their strings corrupted when Encoding.default_internal is set to nil. -
- [Bug #20598]
Just like [Bug #20595], Encoding#name_list and Encoding#aliases can have
their strings corrupted when Encoding.default_internal is set to nil.
Co-authored-by: Matthew Valentine-House <matt@eightbitraptor.com> -
04:00 PM Revision 9c5e9d29 (git): [ruby/openssl] rewriting most of the asn1 init code in ruby
- to have as much of the lib in ruby as possible
https://github.com/ruby/openssl/commit/8305051728 -
02:53 PM Bug #20597 (Closed): `eval('break if false')` should raise SyntaxError but retuns nil
- These are all SyntaxError (Invalid break, compile error (SyntaxError))
~~~ruby
ruby -ce "break if false"
ruby -ce "break if (false)"
ruby -ce "break if nil"
ruby -ce "break if (nil)"
ruby -ce "break if 0>1"
~~~
But when it is p... -
- Fix corruption of encoding name string
[Bug #20595]
enc_set_default_encoding will free the C string if the encoding is nil,
but the C string can be used by the encoding name string. This will cause
the encoding name string to be corrup... -
02:03 PM Bug #20596 (Closed): [BUG] unexpected rb_parser_ary_data_type (0) for script lines
- https://github.com/ruby/setup-ruby/actions/runs/9697443755/job/26761634004
```
/home/runner/.rubies/ruby-head/bin/bundle install --jobs 4
/home/runner/.rubies/ruby-head/lib/ruby/3.4.0+0/rubygems/vendor/molinillo/lib/molinillo/dependen... -
- Applied in changeset commit:git|c6a0d03649c686a537c1f513a1e32205ac6a6512.
----------
Fix corruption of encoding name string
[Bug #20595]
enc_set_default_encoding will free the C string if the encoding is nil,
but the C string can be u... -
- [Bug #20595]
enc_set_default_encoding will free the C string if the encoding is nil,
but the C string can be used by the encoding name string. This will cause
the encoding name string to be corrupted.
Consider the following code:
... -
01:44 PM Misc #20013: Travis CI status
- As a note, by the Travis's above change, while people can run Travis by sending a pull-request to the ruby/* repositories, people cannot run Travis by pushing commits to the branches in people's forked repositories of the ruby/* reposito...
-
- > I am working to get the list of the github accounts who contributed to the ruby/ruby, ruby/zlib and ruby/prism in the last 2 years, and submit the list to Travis support to enable Travis for their forked repositories.
I am working t... -
- https://github.com/rubygems/rubygems/commit/979cd898f9
-
- https://github.com/rubygems/rubygems/commit/2954ab79a9
-
- RubyGems >= 3.5 no longer raises `Gem::FilePermissionError` explicitly.
https://github.com/rubygems/rubygems/commit/df54b9fd90 -
02:44 AM Revision 0c7776a2 (git): [DOC] Fix a link in NEWS.md
06/26/2024
-
09:59 PM Misc #20574: DevMeeting-2024-07-11
- * [Feature #19979] Allow methods to declare that they don't accept a block via `&nil`
* Based on the discussion in DevMeeting at RubyKaigi 2024, it seems this feature is still considered valuable.
* I've rebased @nobu 's [original ... -
- The result of the discussion at the face to face dev meeting in Naha was that while #15554 is preferred, it has some drawbacks and an explicit method for declaring that a method accepts no blocks is still a good thing to have.
For th... -
08:46 PM Revision ce7299d0 (git): [ruby/prism] Handle block exits under modifiers
- https://github.com/ruby/prism/commit/6b78f5309b
-
- The Ruby 3.3 backport PR is here: https://github.com/ruby/ruby/pull/11063
-
- GitHub PR: https://github.com/ruby/ruby/pull/11062
enc_set_default_encoding will free the C string if the encoding is nil, but the C string can be used by the encoding name string. This will cause the encoding name string to be corrup... -
08:01 PM Revision 4cbc41d5 (git): [YJIT] Fix block and splat handling when forwarding
- This commit fixes splat and block handling when calling in to a
forwarding iseq. In the case of a splat we need to avoid expanding the
array to the stack. We need to also ensure the CI write is flushed to
the SP, otherwise it's possibl... -
05:31 PM Revision 43d7db38 (git): [DOC] Doc for exceptions (#11008)
-
- https://github.com/rubygems/rubygems/commit/39a47c264e
-
03:51 PM Feature #20594: A new String method to append bytes while preserving encoding
- > I consulted @maximecb (Maxime Chevalier-Boisvert) about this proposal, and according to her, accepting variable arguments makes it harder for YJIT to optimize.
Yeah my personal preference would be for something like `bytepush` or `b... -
01:39 PM Revision aa2a88ce (git): Use `File.open` instead of `IO.sysopen` (backport to 3.5) (#11061)
- Use backport to 3.5
-
11:24 AM Revision c6c817df (git): [rubygems/rubygems] Add --no-test, --no-ci, and --no-linter options
- https://github.com/rubygems/rubygems/commit/f58660ffcc
-
10:01 AM Revision 04c86a63 (git): [ruby/date] Update zonetab.h at 2024-06-26
- https://github.com/ruby/date/commit/ef5a0dac5b
-
05:45 AM Revision b4503d79 (git): Mentioned new mswin tasks related with vcpkg
-
- 04:55 AM Revision f63b54ca (git): Update bundled gems list as of 2024-06-26
-
- Use rbs#1912
-
-
06/25/2024
-
10:48 PM Revision 4fb7e1b6 (git): Change `enum rb_parser_ary_data_type` default value to 1 for easy debug
- We face `[BUG] unexpected rb_parser_ary_data_type (0) for script lines`
on master branch recently.
This commit changes `enum rb_parser_ary_data_type` to start with `1`
and `0` to be invalid then it makes clear `rb_parser_ary_data_type (0... -
08:15 PM Revision a47c152a (git): [ruby/prism] Fix documentation comment in generated with templates files
- Fix warning
```
/******************************************************************************/
^
warning: [dangling-doc-comments] documentation comment is not attached to any declaration
```
in Nodes.java and AbstractNodeVisitor.java ... -
- +1 from me.
I forgot that `String#concat` accepted integers.
Together with accepting variable arguments, it feels like this new method would be quite a bit complicated to implement and optimize well.
But, it's consistent with `Strin... -
- ### Context
When working with binary protocols such as `protobuf` or `MessagePack`, you may often need to assemble multiple
strings of different encoding:
```ruby
Post = Struct.new(:title, :body) do
def serialize(buf)
buf... -
02:32 PM Revision b88ac94e (git): [rubygems/rubygems] Only validate resolution info in Bundler
- This commit switches out the full gemspec validation for a partial one
which only performs resolution related checks. This will allow gem
authors to run `bundle` commands immediately after creating a new gem
with Bundler, rather than hav... -
- This method validates only what is required for resolution, skipping any
irrelevant metadata validation. This will be used by Bundler instead of
doing a full validation, allowing gem authors to use `bundle` commands
immediately in newly ... -
-
12:55 PM Revision bfb4272d (git): [rubygems/rubygems] Regenerate bundler docs for June 2024.
- https://github.com/rubygems/rubygems/commit/72103ca1e8
-
-
- 06:59 AM Revision 501b22aa (git): Update bundled gems list as of 2024-06-25
-
- * [Bug #20593] `Kernel#format` emits a `too many arguments for format string` warning when called with a single hash and no key is used
* `format` emits warnings when too many positional arguments were passed, but doesn't do so if too... -
-
-
-
- Is there the criteria for reverting this until Ruby 3.4 release?
-
- Applied in changeset commit:git|250fc1223c4dc293f8759ddedce9cba7841f1474.
----------
[Bug #20457] Do not remove final `return` node
This was an optimization for versions prior to 1.9 that traverse the
AST at runtime. -
- This was an optimization for versions prior to 1.9 that traverse the
AST at runtime. -
-
- (https://github.com/ruby/reline/pull/727)
https://github.com/ruby/reline/commit/be45660c83
06/24/2024
-
06:33 PM Bug #20043 (Closed): `defined?` checks for method existence but only sometimes
- Applied in changeset commit:git|ae0c7faa79029ebe615487d23c8ee1ca44ce4a36.
----------
Handle hash and splat nodes in defined?
This supports the nodes in both in the parse.y and prism compilers.
Fixes [Bug #20043]
Co-authored-by: Kevin... -
- This supports the nodes in both in the parse.y and prism compilers.
Fixes [Bug #20043]
Co-authored-by: Kevin Newton <kddnewton@gmail.com> -
- https://github.com/ruby/io-console/commit/4b2e876dd7
-
- https://github.com/ruby/io-console/commit/f8ec8a0617
-
- https://github.com/ruby/io-console/commit/3798aae42d
-
- https://github.com/ruby/date/commit/cfbd6a6b13
-
- https://github.com/ruby/date/commit/5974ac9c7e
-
- https://github.com/ruby/date/commit/5d67437b1f
-
-
- Proposed patch: https://github.com/ruby/ruby/pull/11053
-
02:13 PM Bug #20593: `Kernel#format` emits a `too many arguments for format string` warning when called with a single hash and no key is used
- Huh, interesting. I didn't realise it worked that way. I recently fixed a few of those in rubocop but didn't catch on that it only applies when the string takes no arguments.
This makes sense to me. Fixing these warnings didn't improv... -
- `Kernel#format` emits a warning when too many arguments are passed.
It does so for postional arguments, which make sense, as it could be indicative of a bug:
```ruby
$VERBOSE = true
format("%s", 1, 2) # warning: too many argum... -
- [Feature #20205]
This was an undesired side effect. Now that this value is a triplet, we can't
assume it's disabled by default. -
- I found a small repro for that last one:
```ruby
class Model
class << self
def scope(name, body, &block)
singleton_class.define_method(name) do |*args|
new._exec_scope(*args, &body)
end
singleton... -
- Thanks for the update. The reason why I added another failure is to record which commit triggers the failure. No rush.
-
- There is also:
```
Assertion Failed: /usr/src/ruby/include/ruby/internal/core/rarray.h:257:rb_array_len:RB_TYPE_P(a, RUBY_T_ARRAY)
ruby 3.4.0dev (2024-06-23T06:59:37Z master acce96647d) +YJIT [x86_64-linux]
actual type: 1-- Control... -
- So that `proxy_controller_test` failure I can't repro locally it seems, but I see it actually crashed on another build with YJIT in debug mode:
```
Assertion Failed: /usr/src/ruby/vm_callinfo.h:115:vm_ci_packed_p:imemo_type_p((VALUE)... -
- I'll try to get the underlying error for that one.
@tenderlovemaking is looking at these issues, but he's at a company event this week, so not much will happen until next week. -
- There is another Rails CI failure https://buildkite.com/rails/rails-nightly/builds/683#01903773-e527-4777-8dd1-7656c1ab9cbc/1214-1277 since cdf33ed5f3
```
$ ruby -v
ruby 3.4.0dev (2024-06-18T16:28:25Z master cdf33ed5f3) [x86_64-linu... -
09:14 AM Revision 9cfc1362 (git): Extend the timeout of test_try_ldflag_invalid_opt and test_try_cppflag_invalid_opt
-
- Since signal handlers just set flag and return now, `sigsetjmp` and
`siglongjmp` will not be needed. -
07:50 AM Revision 92f2cf1b (git): With ASAN `EC_REPUSH_TAG` doesn't work
- With ASAN, `EC_TMPPOP_TAG`/`EC_REPUSH_TAG` doesn't work
so rewrite it without them. They are used for the performance,
but here calling stacked `END{}` and not so performance critical. -
04:25 AM Revision afd42e4d (git): Use $GITHUB_PATH
-
-
- GitHub Actions macos-arm-oss is often too slow and does not timestamp as
expected. -
-
-
-
-
-
12:03 AM Revision 18404af9 (git): .github/actions/setup/macos/action.yml - use brew openssl@3, not 1.1
06/23/2024
-
04:58 PM Feature #18762: Add an Array#undigits that compliments Integer#digits
- I'll speak up for this feature. Whilst writing a base58 encoder/decoder today, I was surprised by the lack of an inverse to `Integer#digits`.
In my benchmarking the reduce/inject variant was considerably faster than the sum variant, esp... - 06:59 AM Revision acce9664 (git): Update bundled gems list as of 2024-06-22
-
06/22/2024
-
- Thanks! Looks good to me. I have merged your PR.
-
01:05 PM Bug #20592 (Closed): Interrupting Addrinfo causes Segmentation fault on alpine
- Applied in changeset commit:git|fba8aff7af450e476e97b62385427dfa51850955.
----------
[Bug #20592] Fix segfault when sending NULL to freeaddrinfo
On alpine freeaddrinfo does not accept NULL pointer -
- PR https://github.com/ruby/ruby/pull/11045
-
- On Alpine 3.20 got a segfault when interrupting Addrinfo request which would resolve with error.
Same on
ruby 3.3.1 (2024-04-23 revision c56cd86388) [x86_64-linux-musl]
Tried on docker image alpine3.20 and ruby:3.4.0-preview1-alp... -
- On alpine freeaddrinfo does not accept NULL pointer
-
-
06:12 AM Revision 73b7eebf (git): build: fix crossruby build by handling empty `ac_abs_builddir`
- `ac_abs_builddir` can be empty when the build is top-level (not
subdirs, and Ruby is usually the case). In such case, the MINIRUBY is
expanded to `$RUBY -I -r'$(arch)-fake'`, which interprets `-r$(arch)-fake`
as an argument to `-I` optio... -
- Also executing variable containing an option may not be portable.
Follow up of dd378c5a2483002d50053cf81d00004c1fb3c8bd. -
- Currently, the macos.yml file used for GHA CI is using OpenSSL 1.1.1, and the same for the builds done in https://github.com/ruby/ruby-dev-builder.
But, all the macOS builds of 3.1 thru 3.3 are using OpenSSL 3.x?
See the 'Ruby Info...