Activity
From 09/17/2026 to 09/23/2026
Today
-
05:53 AM Revision 852b2434 (git): Fix use-after-free in String#unpack
- If the source string is modified in the block of String#unpack, it can
cause an use-after-free. For example, the following script crashes:
fmt = "A" * 1_000_000
str = "C" * 1_000_000
str.unpack(fmt) { fmt.clear } -
05:36 AM Bug #22377 (Closed): compile.c: `getlocal` not turned into `getlocal_WC_0`
- Applied in changeset commit:git|0fab30f723810af1608d2162e2aef32b44a1ee3a.
----------
compile.c: ensure swapped element isn't skipped
[Bug #22377]
When we statically resolve a `swap` we need to make sure the element
we reordered will r... -
- Thanks for the repro, I think I know what's going on, I'll look into it.
-
- [Bug #22377]
When we statically resolve a `swap` we need to make sure the element
we reordered will receive its optimization pass. -
05:32 AM Revision 592f1ffd (git): Bump v3.4.11.
-
- ruby_3_4 commit:7bbe8e88007f8283a421cb31db86da32076126a6 merged revision(s) commit:ad23fd4a519fde4fc96c7326fb9b0347cbc347b8.
-
- ruby_3_4 commit:41e214d7100100a43c469e2628403d6916d1f0f8 merged revision(s) commit:d87ffb271a80fb108de3465d12ada829f3c130af.
-
- ruby_3_4 commit:75309143a8494abd682bdd23f3b18ecc2e48a2e4 merged revision(s) commit:cb302881629f997f403e705425f69e5f6b0741ac, commit:c914aeb622ce7337223cd5c23f30daf9f7d3e7c4.
-
- ruby_3_4 commit:93d1c2d7c3782cfbe0448849bd2cb79ce66c3860 merged revision(s) commit:ee786598f10c6def72265fb7105fe505f563f52f, commit:54f254aa41eb624429e6b884c0a6381eeeac50d0, commit:15dbb08ba8b69877c8012e0f5950f3f26a3abffe, commit:249de20...
-
- MEMO: related commit hashes.
commit:ee786598f10c6def72265fb7105fe505f563f52f
commit:54f254aa41eb624429e6b884c0a6381eeeac50d0
commit:15dbb08ba8b69877c8012e0f5950f3f26a3abffe
commit:249de200d5be1f1d1f936e40949a6b48ac9b6e25 -
- [PATCH] [Bug] win32: create directory symlinks for relative directory targets
File.symlink decided whether to create a directory symlink by calling
GetFileAttributesW on the target as given, which resolves a relative
... -
- [PATCH] [Bug #22335] Use identity hashes for keyword argument tracking
With more than 31 keyword arguments, tracking unevaluated defaults
uses an integer-keyed Hash whose collisions can invoke methods when
`Integ... -
- [PATCH] [Bug #22340] Fix stat on symbolic-link loops on Windows
Preserve `ELOOP` instead of hiding symbolic-link resolution errors
with the file-enumeration fallback, so existence checks return false. -
- [PATCH] Fix memory leak of transcoding when fallback raises
When the fallback function in transcode_loop raises, it will leak the memory
in rb_econv_t. The following script reproduces the leak:
10.times do
... -
04:05 AM Revision 27be6ccd (git): ZJIT: Avoid creating IseqPayload on rb_zjit_iseq_free (#18953)
- 02:41 AM Revision 8c0eb987 (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 2.87.16 to 2.87.17
- [Release notes](https://github.com/taiki-e... -
- The precision in Time.new calls to_int which can modify the string passed
into Time.new. This can cause an use-after-free. For example, the following
script crashes:
str = "2000-01-01 00:00:00" + "0" * 1_000_000
obj = Object.new... -
12:03 AM Revision b20a1ab6 (git): [DOC] Harmonize file size methods
-
-
09/22/2026
- 11:20 PM Revision 59bd0509 (git): ZJIT: Change HasType into CondBranchHasType (#18633)
- We use `HasType`+`CondBranch` a lot—it is how we do polymorphic dispatch in HIR—so it's pretty hot. Without smarter instruction fusion/instruction selection, this generates a lot of code. Kokubun identified this in https://github.com/Sho...
-
11:17 PM Bug #22377 (Closed): compile.c: `getlocal` not turned into `getlocal_WC_0`
- Missed by the "operands_unification" pass, somehow.
See 0023 below:
```text
$ ./miniruby --parser=parse.y --dump=insns -e '
assign = ->(outer) {
m = Module.new
m.set_temporar... - 10:15 PM Revision dcbf89e0 (git): ZJIT: Make LIR Abort a terminator (#18967)
- It's an abrupt end to a block, but it definitely ends the block. No need
to put a return instruction after it. -
09:37 PM Feature #22376 (Open): Undefine the allocator for Class
- Uninitialized classes, created by `Class.allocate`, are a recurring source of bugs (most recently #22341). Every place that inspects a class's superclass chain has to check whether the class has been initialized, and it's easy to forget ...
- 09:18 PM Revision 7e0d49ae (git): ZJIT: Add HIR comment to terminators too (#18890)
- We previously only had this in `gen_insn` which left terminators
un-commented. -
08:10 PM Bug #22375 (Open): `Time.new(String)` can disagree with itself about the date
- `Time.new("2026-02-30T00:00:00Z")` returns a `Time` object that is internally inconsistent.
It shows as Feb 30 or Mar 2 depending on what methods are called:
```ruby
t = Time.new("2026-02-30T00:00:00Z") # => 2026-02-30 00:00:00 UTC
... - 07:52 PM Revision e77b7b6f (git): ZJIT: Specialize Struct class .new (#18835)
- We can use `ObjectAllocClass` for Structs because they are basically
a `__members__` ivar lookup for sizing plus the default allocator.
Requires #18888 to make sure the constructor does not raise. -
07:06 PM Revision bf4ab046 (git): [ruby/prism] Move PRISM_NODISCARD to the correct position, again
- (https://github.com/ruby/prism/pull/4232)
Followup to https://github.com/ruby/prism/commit/ba16ae22561179998e4208e4c37262b47f693c11
I can't run rake `cargo:test`:
```
$ cargo build
Compiling ruby-prism-sys v1.9.0 (ruby-prism/rust/ru... -
- It might be worth noting for anyone that wants to try this, if you do LTO with fat LTO objects and a static build (as described at the top) this will make native extension compilation much slower.
You can speed this up by stripping t... - 04:36 PM Revision 8d4efc14 (git): ZJIT: Elide Kernel#dup for more classes (#18872)
- `rb_obj_dup` skips immediates:
```c
VALUE
rb_obj_dup(VALUE obj)
{
VALUE dup;
if (special_object_p(obj)) {
return obj;
}
// ...
}
static inline int
special_object_p(VALUE obj)
{
if (SPECIAL_CONST_P(obj)) re... -
04:22 PM Bug #22342: IO#gets can ignore the limit argument in some edge cases
- I've opened a GitHub PR for this: https://github.com/ruby/ruby/pull/18961
-
- `io.gets(rs, limit)` ignores the `limit` value and reads until `rs` is found or EOF, if the `limit`-th byte of the stream matches `rs`'s last byte and:
1. The potential start of the record separator, the `limit-rs.bytesize`-th byte of... - 03:08 PM Revision 720994eb (git): ZJIT: Fold Class#superclass (#18682)
- I saw that YJIT has a fast-path for this so I figured we should too.
-
02:58 PM Revision cec5c97b (git): [ruby/fileutils] Add tests for `FileUtils.options_of`
- Fix https://github.com/ruby/fileutils/pull/195
https://github.com/ruby/fileutils/commit/b2d9b72739 -
- Fix https://github.com/ruby/fileutils/pull/194
https://github.com/ruby/fileutils/commit/8b1efe473e -
- Fix https://github.com/ruby/fileutils/pull/193
https://github.com/ruby/fileutils/commit/6fd04f3cea -
- Fix https://github.com/ruby/fileutils/pull/192
https://github.com/ruby/fileutils/commit/df83e4bb04 -
- Fix https://github.com/ruby/fileutils/pull/191
https://github.com/ruby/fileutils/commit/ff78cfd3a4 -
- Windows reports self-referential links as existing, so use a missing
target to test a dangling link rather than a symbolic-link loop.
https://github.com/ruby/fileutils/commit/803fe3be07 -
- Cover single, multiple, and dangling links through `ln_s`, keeping
valid-link tests available where dangling links are unsupported.
Fix https://github.com/ruby/fileutils/pull/190
https://github.com/ruby/fileutils/commit/189aefb679 -
- Cover removal of existing and missing files through `rm_f` and
verify the `safe_unlink` alias.
Fix https://github.com/ruby/fileutils/pull/189
https://github.com/ruby/fileutils/commit/902bcce36b -
- Cover removal of existing and missing trees through `rm_rf` and
verify the `rmtree` alias.
Fix https://github.com/ruby/fileutils/pull/188
https://github.com/ruby/fileutils/commit/07cc858f16 -
- Check the result and effects of `mv` and verify the `move` alias.
Fix https://github.com/ruby/fileutils/pull/187
https://github.com/ruby/fileutils/commit/3bed6e74a1 -
- Check `mkdir_p` return values and verify the `makedirs` and `mkpath`
aliases without duplicating directory creation tests.
Fix https://github.com/ruby/fileutils/pull/186
https://github.com/ruby/fileutils/commit/72ff748e26 -
- Verify that `ln` creates hard links and that `link` remains an alias.
Fix https://github.com/ruby/fileutils/pull/185
https://github.com/ruby/fileutils/commit/46db9d557b -
- Avoid duplicating working-directory coverage already in `test_pwd`.
Fix https://github.com/ruby/fileutils/pull/183
https://github.com/ruby/fileutils/commit/d937a3d5ba -
- Cover directory changes, block behavior, and verbose output through
`cd`, and verify that `chdir` remains an alias.
Fix https://github.com/ruby/fileutils/pull/182
https://github.com/ruby/fileutils/commit/8476509760 -
- Fix https://github.com/ruby/fileutils/pull/181
https://github.com/ruby/fileutils/commit/4b041318e9 -
- Cover equal and different file contents in `test_cmp` and verify
that `compare_file` and `identical?` remain aliases.
Fix https://github.com/ruby/fileutils/pull/180
Fix https://github.com/ruby/fileutils/pull/184
https://github.com/ruby... -
- Great catch! This is my mistake from https://github.com/ruby/ruby/commit/e01e89f55c82be8db5d8ccbf305ee38e3769e582
It should be safe to remove that assertion, but I also think it should probably be forbidden to include on an uninitiali... -
02:20 PM Bug #22341 (Open): Class#superclass can crash the interpreter in debug build
- While implementing this in ZJIT, @nirvdrum reported, and I investigated, a potential interpreter crash:
module M; end
C = Class.allocate
C.include M
C.superclass
In a debug build, this crashes:
plum% ./bui... -
- Applied in changeset commit:git|ad23fd4a519fde4fc96c7326fb9b0347cbc347b8.
----------
[Bug #22340] Fix stat on symbolic-link loops on Windows
Preserve `ELOOP` instead of hiding symbolic-link resolution errors
with the file-enumeration f... -
- On Windows, `File.stat` succeeds on a self-referential symbolic link instead of raising `Errno::ELOOP`, and `File.exist?` returns `true`. Opening the same link raises `Errno::ELOOP`.
I noticed this while adding symbolic-link tests to... -
- Preserve `ELOOP` instead of hiding symbolic-link resolution errors
with the file-enumeration fallback, so existence checks return false. - 01:48 PM Revision 7ba699ae (git): Make fibers RUBY_TYPED_THREAD_SAFE_FREE
- Now fibers can be collected by multiple ractors at once when those Ractors
are running local GCs. Fibers are created in the Ractor that owns them, so
if multiple Ractors have unmarked fibers during local GC sweeping, we can
now collect t... - 01:39 PM Revision 1070504b (git): Make non-Array Struct __members__ a bug (#18888)
- We should not ever hit this from Ruby code; user code cannot modify
ivars that do not start with @.
This makes `num_members` leaf and therefore Struct class allocation also
leaf. - 01:36 PM Revision ce75c108 (git): Allow disabling `zjit-stats` in tool/zjit_diff.rb (#18893)
- When running benchmarks, sometimes there is a overhead in the statistics collection, so it can be useful to disable `zjit-stats`.
In my case I was investigating https://github.com/Shopify/ruby/issues/1051 and was confused why I was not ... -
- [PATCH] Fix out-of-bounds in Array#permutation when array modified
Calling to_int on the count may run Ruby code that modifies the array.
This can cause an out-of-bounds in Array#permutation. For example, the
fol... -
- [PATCH] Fix buffer overflow in String#rpartition
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
The conversion of the pattern to string can cause the string to b... -
- If the Dir object is closed in the block of Dir#each/each_child/scan, then
it will crash because dirp->dir will be a NULL pointer. The following
script demonstrates the crash:
d = Dir.open("/")
d.each { d.close } -
-
- Also simplify the notation of the special const reserved words.
-
- Show that dangling symbolic links return false and clean up the
example files.
Follow up on ruby/ruby#18571. -
- Calling to_int on the count may run Ruby code that modifies the array.
This can cause an out-of-bounds in Array#permutation. For example, the
following code crashes:
ary = (1..1_000).to_a
obj = Object.new
obj.define_singleto... -
- 02:11 AM Revision 6454569c (git): Bump the github-actions group across 1 directory with 5 updates
- Bumps the github-actions group with 5 updates in the / directory:
| Package | From | To |
| --- | --- | --- |
| [ruby/setup-ruby](https://github.com/ruby/setup-ruby) | `1.323.0` | `1.324.0` |
| [github/codeql-action/init](https://github...
09/21/2026
-
- The conversion of the pattern to string can cause the string to be modified.
This can cause a buffer overflow since it uses the original length of the
string. For example, the following script reports a buffer overflow with
ASAN:
s ... -
- https://github.com/ruby/mmtk/commit/3f94f904eb
-
04:46 PM Bug #22339 (Open): Ruby::Box: crash / hang at VM shutdown with RUBY_FREE_AT_EXIT=1
- ## Summary
With `RUBY_BOX=1` and `RUBY_FREE_AT_EXIT=1`, Ruby crashes or hangs while freeing boxes at VM
shutdown. No user box needs to be created: the built-in boxes are enough.
`box_entry_free()` walks `box->classext_cow_classes`... -
- Fix ruby/rdoc#1592.
-
03:58 PM Revision f07f65b7 (git): [PRISM] Source file no longer static literal
-
- Ref: https://bugs.ruby-lang.org/issues/17593.
Because `__FILE__` is currently considered a static literal,
it's entirely evaluated during compilation.
The problem is that ISeqs can be serialized and reloaded later,
but since `__FILE__`... -
02:33 PM Bug #22335: Stack corruption with >31 keyword args and refined Integer#==
- Thank you for the prompt fix.
Not an area of ruby I'm very familiar with, but I think perhaps we should also be extending sp in `vm_callee_setup_arg` and `vm_call_iseq_setup_kwparm_kwarg`? Getting rid of the method dispatch fixes the ... - 01:55 PM Revision 61de3dd7 (git): [ruby/openssl] Make ruby/openssl types RUBY_TYPED_THREAD_SAFE_FREE
- It's safe to free these types during Ractor-local GC in Ruby 4.1. We've
seen the freeing of this type show up in profiles before.
https://github.com/ruby/openssl/commit/e481c6408f
Co-authored-by: Kazuki Yamaguchi <k@rhe.jp> -
12:40 PM Feature #22068: Adding post-quantum cryptography (PQC) support across Ruby standard libraries
- Note that I sent PR to support SLH-DSA in ruby/openssl. However, I don't have a plan to add the SLH-DSA tests across the standard libraries except for ruby/openssl. Because I think SLH-DSA is a kind of unpopular PQC algorithm used for th...
-
- I agree that this is worth backporting, I did https://github.com/ruby/ruby/pull/18945 for 4.0. Backport for 3.4 is fine too I guess, it someone wants to do it.
-
- We need to keep track of pending pages in RubyHeapTrigger otherwise a large
allocation may cause infinite number of GCs to be ran since it will always
appear that we have enough space in the heap (and thus the heap won't grow)
but still ... -
- @jneen you should probably open this as a PR to ruby, yes. The pr to upstream can happen after.
Currently ruby has no tests for duplicate character classes I think, so it would be nice to add some tests in that PR to show it still war... -
09:14 AM Bug #22325 (Closed): out of bounds read in array
- Applied in changeset commit:git|8befacf9e6c898a0f79dcb53bdd2e4cea24e0d1c.
----------
Guard against oob read in rb_ary_aref1
When the array is shrunk during rb_arithmetic_sequence_beg_len_step,
ary_subseq_len returns -1. Return an empty... -
- When the array is shrunk during rb_arithmetic_sequence_beg_len_step,
ary_subseq_len returns -1. Return an empty array instead of passing the
negative length to ary_make_partial or ary_make_partial_step.
[Bug #22325] -
- Fix https://github.com/ruby/fileutils/pull/177
https://github.com/ruby/fileutils/commit/86278df2f7 -
- Fixed by ruby/ruby#17545.
https://github.com/ruby/fileutils/commit/9e957c2330 -
- [PATCH] Fix use-after-free in String#% when array modified
If the array object passed into String#% is modified during conversion
(such as via to_s, to_i, to_f, etc.) then it can be a use-after-free if
the buffer... -
- [PATCH] Fix out-of-bounds access in String#byteindex/byterindex
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
If the source string is modified when converting o... -
- vo.x (Vit Ondruch) wrote:
> My proposal is to change the `require_relative` in following way:
> ...
Do you mean `require_relative` should load the file in the current working directory?
I don't think it is reasonable behavior.
`requi... -
- https://github.com/ruby/ruby/pull/17545
-
- If the source string is modified when converting offset to an integer,
then there could be an out-of-bounds access because the length of the
string is captured before to_int is called. For example, the following
script triggers an ASAN e... -
05:31 AM Feature #22274: Make `IO::Buffer` no longer experimental.
- > Thanks, that makes sense. At the moment a locked buffer can be frozen, and #locked on a frozen buffer still increments lock_count, so I assume both will be handled as part of the implementation.
Correct :) -
05:26 AM Feature #22274: Make `IO::Buffer` no longer experimental.
- ioquatix (Samuel Williams) wrote in #note-10:
> As per my design, a buffer that is frozen does not need to update or touch `locked_count` (in fact, only a buffer with `locked_count == 0` can be frozen). Once frozen, all operations (that... -
- ioquatix (Samuel Williams) wrote in #note-6:
> `freeze` is for "safe to share this Ruby object" and `readonly` means the backing buffer cannot be modified by this view.
As far as I know, there is currently no API to make an existing ... -
- > Even on a frozen buffer, #locked updates lock_count with a non-atomic read-modify-write, so two Ractors calling #locked on the same buffer would cause a data race on it. It seems that freezing alone is not sufficient for sharing.
As p... -
- ioquatix (Samuel Williams) wrote in #note-6:
> Therefore, I'm against a frozen buffer meaning underlying memory cannot be modified.
If Matz agrees, I don't have a strong opinion on how freeze should be interpreted. However, under thi... -
- ioquatix (Samuel Williams) wrote in #note-6:
> For the purpose of sharing `IO::Buffer`, `freeze`ing this state is sufficient.
Even on a frozen buffer, `#locked` updates `lock_count` with a non-atomic read-modify-write, so two Ractors... - 03:07 AM Revision 647e6e61 (git): [ruby/fileutils] Remove redundant condition
- Since `Process.uid` returns `0` always on mingw or mswin, `root_in_posix?` returns `true` too.
https://github.com/ruby/fileutils/commit/d18fbf0c56 - 03:07 AM Revision be820c30 (git): [ruby/fileutils] Skip test_rm_r_no_permissions test under root
- Skip the test_rm_r_no_permissions test under the root user, as deletion always succeeds.
Signed-off-by: Jiaying Song <jiaying.song.cn@windriver.com>
https://github.com/ruby/fileutils/commit/3c831389c5 -
03:05 AM Revision 508ee642 (git): [ruby/fileutils] Skip chown tests when the target group is not assignable
- In user-namespace environments (e.g. ChromeOS Crostini) getgroups(2) can report supplementary groups such as the overflow GID 65534 (nobody) that a non-root process cannot actually chgrp a file to. TestFileUtils#setup built @groups strai...
- 02:59 AM Revision 0cb17e33 (git): [ruby/fileutils] Remove unused fu_copy_stream0 helper
- fu_copy_stream0 has had no caller since b26846bf replaced both uses with direct IO.copy_stream calls in 2008. The retained private wrapper is unreachable.
https://github.com/ruby/fileutils/commit/feba38f651 -
02:57 AM Revision 26f43967 (git): [ruby/fileutils] Fix relative symlink target generation
- https://github.com/ruby/fileutils/commit/7de921e801
-
- String#tr can crash if keys are removed from the translation hash since
we stack allocate a buffer pairs. If entries are deleted during runtime,
then we won't fill the pairs buffer which can crash because it will be
reading uninitialized... -
-
-
-
-
-
01:25 AM Feature #15833 (Rejected): Some refactors for shared-root array
-
- This ticket is too old
09/20/2026
-
11:50 PM Feature #22304: Add rb_warn_to_remove_at() for deprecation warnings shown by default
- Dan0042 (Daniel DeLorme) wrote in #note-8:
> Does the updated target version need to be backported to every branch ?
There is a similar case. Ruby 3.3 still says "Process::Status#& is deprecated and will be removed in Ruby 3.4", althoug... -
05:17 PM Feature #22304: Add rb_warn_to_remove_at() for deprecation warnings shown by default
- > A major version bump is a chance to introduce incompatibilities, so I think we should decide for each deprecation whether to keep the schedule or to remove it at the major version.
Ok I see, that makes sense. It might also be useful... -
- ioquatix (Samuel Williams) wrote in #note-6:
> Implemented in [https://github.com/ruby/ruby/pull/18911](https://github.com/ruby/ruby/pull/18911)
Great, this will make semantics match more closely and ensure reliable behavior.
ioquatix ... -
-
- Once all `*.rbinc` files have been generated, the command should not
be required anymore. -
- If the array object passed into String#% is modified during conversion
(such as via to_s, to_i, to_f, etc.) then it can be a use-after-free if
the buffer of the array is freed.
This commit changes it to detect if the buffer has been cha... -
- Avoid incorrect variadic argument forwarding in `RUBY_ASSERT_MESG_WHEN`
by the traditional MSVC preprocessor, which breaks `VM_ASSERT` when
`RUBY_DEBUG` is enabled. -
- [PATCH] Raise the ModGC `make check` timeout for MMTk to 90 minutes
MMTk's `make check` on macos-26 takes 27 minutes at the median and 37 at the 90th percentile, so the default 40 minute step limit sits inside the distribution o... -
- [PATCH] Add timeout scale to macOS ModGC CI jobs
Same fix as commit 8e8682fc2312164c49d3bf4a15d1f464d2644598, which added
RUBY_TEST_TIMEOUT_SCALE=10 to all macOS check jobs in macos.yml but did
not cover modgc.ym... -
- ruby_3_4 commit:bce3c89801d667a2cc38940228b8656f46879e7e merged revision(s) commit:b61da996afb02715cb16703a55f57c4417fd2d5b.
-
- ruby_3_4 commit:350ef312531ac79eb2771f6d5cd242d720657d0b merged revision(s) commit:d7fe390c85d1a0aeda104026134d19b86199f9b7.
-
- ruby_3_4 commit:b82e5058907869051d336449dd67f16b007ad7b8 merged revision(s) commit:20ca4e1f647107df1a5ae75b514f20b587091405.
-
- ruby_3_4 commit:2a77c0f6f9ee7752a3594455bac24812d8aeba9d merged revision(s) commit:0c9f3f1e68e23a9ba2d7fdfdb86bd7ddcaf5e9d9.
-
- ruby_3_4 commit:3ebd235fb667e37b9f5cb4dcf61b19f07c4d2b34 merged revision(s) commit:5eeaedc3d3014a12264aa252c751e14ba0fc5507.
-
- ruby_3_4 commit:ce5cca10a5143da2ccea33a53190e39a6b923241 merged revision(s) commit:685144e3d60513ccde1c6cb4bed20597cb860002.
-
- ruby_3_4 commit:cf53d1a7283dbe78830cabd9e72124208d8380d3 merged revision(s) commit:5831948a66412922e1abafd2c38fd6400e420581.
-
- ruby_3_4 commit:9bfc30e701db71c3bfec531422ebea7e2c1cdfb5 merged revision(s) commit:d38952c45e27becbe160f832aea823546899ad2e.
-
- [PATCH] [Bug #22336] Fix use-after-free in `IO::Buffer#get_string`. (#18919)
-
- [PATCH] Fix crash when returning splat
[Bug #22334]
If we return a with a keyword splat, then each element does not necessarily
push two values onto the stack, so we cannot compile this as a simple sequence
... -
- [PATCH] Fix Method#inspect when klass is iclass
[Bug #22321]
In the following script, the klass of the Method object is an iclass, which
will raise an error "NoMethodError: undefined method 'inspect' for an inst... -
- [PATCH] [Bug #22322] Take transcoder names from declared transcoders
Fix use-after-free of dangling pointers. -
- [PATCH] Fix use-after-free when clearing array during zip
[Bug #22319]
Array#zip triggers an use-after-free it does not account for when the call
to rb_check_array_type modifies the source array. The following s... -
- [PATCH] Fix use-after-free when clearing array during flatten
[Bug #22318]
Array#flatten triggers an use-after-free it does not account for when the
call to rb_check_array_type modifies the source array. The fol... -
- [PATCH] Fix crash when source string modified in String#unpack
[Bug #22315]
The following script crashes because there is a use-after-free on the
source string during String#unpack:
s = "C" * 400000... -
- The 'r/R' specifiers are added in 4.1. I suppose we don't have to backport this.
-
- ruby_3_4 commit:72aa33a12683d7e351ca406a4cad45e5f1f98e26 merged revision(s) commit:e5132b162737829f744b2b23908b9d08479db977.
-
- ruby_3_4 commit:1e1e90990c14805ee327630168360d3b6530aab3 merged revision(s) commit:46c287e2ca39da63d5164e2b7227308a41554c9a.
-
- [PATCH] Fix the `intern` VM instruction to be `leaf=false` (also in YJIT, ZJIT)
rb_str_intern can raise EncodingError and call its initializer.
This call can push a Ruby frame.
Fix in the VM, YJIT, and ZJIT. -
- [PATCH] [Bug #22264] Warn for duplicate keys after leading keyword splat
Leading keyword splats must not suppress duplicate literal key
warnings. -
- [PATCH] Fix splicing an array that shares its buffer with a longer array
Arrays sharing one buffer have the same data pointer, so rb_ary_splice()
took `a` in `b = a.dup; b.pop; b.concat(a)` for `b` itself and rebased
... -
- merged to rubu_3_4 at https://github.com/ruby/ruby/commit/c5a0ddafbbe1ad4477e6c9c60612d5b34ac41f01.
-
- The separator can be modified in the block yielded by String#each_line.
When that happens, the pointer can change which can cause an use-after-free.
For example, the following script crashes:
sep = "x" * 1_000_000
s = "a#{sep}b#... -
-
-
-
09/19/2026
-
- Applied in changeset commit:git|d87ffb271a80fb108de3465d12ada829f3c130af.
----------
[Bug #22335] Use identity hashes for keyword argument tracking
With more than 31 keyword arguments, tracking unevaluated defaults
uses an integer-keye... -
-
- With more than 31 keyword arguments, tracking unevaluated defaults
uses an integer-keyed Hash whose collisions can invoke methods when
`Integer#==` is refined or redefined, corrupting argument locals above
`cfp->sp`. Use identity hashes... -
- The selector conversion may run aribtrary Ruby code and modify the source
string. This can cause a use-after-free if it resizes the source string.
For example, the following script causes corruption:
s = "-" * 1000 + "abc" + "-" * 1... -
11:58 AM Feature #17593: load_iseq_eval should override the ISeq path
- Seems there new movement on related changes but I wanted to add some thoughts.
__FILE__ as syntax rather than method eliminates a bunch of side effect behavior that makes __dir__ troublesome:
* It needs access to the caller frame o... -
-
- The frozen check in `OBJ_FREEZE` makes the separate check redundant,
and immediate values are always frozen, so the old assertion is
unreachable. -
- Avoid calling `rb_obj_freeze_inline` for already frozen objects,
including immediate values, to match `rb_obj_freeze` semantics. -
- `rb_freeze_singleton_class` is an implementation detail of
`RB_OBJ_FREEZE`, so deprecate direct calls from extension libraries
before removing its public declaration in Ruby 4.2. -
-
- @matz
## Slices
Implemented in https://github.com/ruby/ruby/pull/18911
## Freeze
> freezing to mean that the object's own state is immutable
I agree with that. An `IO::Buffer`'s own state consists of:
```c
struct rb_io... -
- Applied in changeset commit:git|b61da996afb02715cb16703a55f57c4417fd2d5b.
----------
[Bug #22336] Fix use-after-free in `IO::Buffer#get_string`. (#18919) -
-
- Thank you for this bug report. I believe [this PR](https://github.com/ruby/ruby/pull/18870) should have fixed this bug. I will mark this issue for backport.
09/18/2026
-
- Thank you for the bug report, I believe this has already been fixed [in this PR](https://github.com/ruby/ruby/pull/18884). I have marked this for backport.
-
06:29 PM Bug #22337 (Closed): OOB write in array.pack()
- Hi, I found a case that causes an oob write in array.pack()
PoC:
```
buf = "Z" * 4096
$buf = buf
evil = Object.new
def evil.to_int
$buf.replace("q")
123456789
end
[evil].pack("r", buffer: buf)
```
asan... -
- Thank you for the bug report, I have [a fix here](https://github.com/ruby/ruby/pull/18919).
-
- Hello, a UAF is possible in IO:Buffer.
PoC:
```
$buf = IO::Buffer.new(64)
evil = Object.new
def evil.to_str
$buf.resize(64 * 1024 * 1024)
"UTF-8"
end
$buf.get_string(0, 32, evil)
```
ASAN (truncated):
```
ERROR... -
- Dan0042 (Daniel DeLorme) wrote in #note-6:
> Sorry for the trouble, but could you explain why "we should do anyway"? I might be missing something here.
A major version bump is a chance to introduce incompatibilities, so I think we shoul... -
- shugo (Shugo Maeda) wrote in #note-5:
> a year cannot be converted to a version at compile time.
I don't understand why converting a year to a version at compile time would be needed. We would need to compare `year >= RUBY_RELEASE_YE... -
- Thank you for the idea. I think the API should keep the version numbers, because the warning message has to tell users the version ("will be removed in Ruby 4.3"), and a year cannot be converted to a version at compile time.
A major ver... -
- I very much like the idea, but the issue with version numbers is that major version jumps break linear assumptions.
For example, in 2023 (Ruby 3.3), if we want to deprecate something in 3 years, we might target 3.6. But as Ruby jumped f... -
- I've created a pull request: https://github.com/ruby/ruby/pull/18904
-
- In addition, how about a higher-level macro that takes both versions of the schedule?
```c
rb_warn_scheduled_deprecation(4.2, 4.3, fmt, suggest, ...)
```
Before 4.2 it warns only when `Warning[:deprecated]` is enabled, from 4.2 it warn... -
- Applied in changeset commit:git|d7fe390c85d1a0aeda104026134d19b86199f9b7.
----------
Fix crash when returning splat
[Bug #22334]
If we return a with a keyword splat, then each element does not necessarily
push two values onto the stac... -
- Thank you for the bug report, I have [a fix here](https://github.com/ruby/ruby/pull/18909).
-
04:27 AM Bug #22334 (Closed): Assertion Failed: VM_ENV_FLAGS:FIXNUM_P(flags)
- The following code:
```ruby
def r; return **[nil]; end; r()
```
Resulted in this output:
```
../vm_core.h:1554: Assertion Failed: VM_ENV_FLAGS:FIXNUM_P(flags)
ruby 4.1.0dev (2026-09-15T05:36:12Z master 3296d5c99c) +PRISM [x86_... -
- [Bug #22334]
If we return a with a keyword splat, then each element does not necessarily
push two values onto the stack, so we cannot compile this as a simple sequence
of key-value pairs combined with newhash. The following script repro... - 09:53 PM Revision 9f6f03ec (git): Make sure VM lock is acquired during GC when single_objspace_p()
- Otherwise we can get assertion failures like:
```
TestGc#test_finalizer_not_run_with_vm_lock [/tmp/_actions-runner-working-dir/ruby/ruby/src/test/ruby/test_gc.rb:1055]:
pid 225354 killed by SIGABRT (signal 6)
| ../src/vm_sync.c:40: ... -
- I'll take a look at seeing if it's the intended change, but let me make
the CI green in the meantime. -
- @matheusrich I considered a keyword argument, but went with positional for consistency.
This isn't yet in a released version, so feel free to come forward with a proposal to make these keyword arguments before the final 4.1 release.
... -
03:17 PM Feature #22297: Module#method_defined? should have an `include_private` argument
- I dislike positional boolean args (see [#17938](https://bugs.ruby-lang.org/issues/17938)). I'm not aware of the performance impact, but would it be possible to use a kwarg here?
-
- Previously, the loop condition checked the wrong direction so required
wrapping around the address space to break out the loop. It takes
a few seconds to do a few million mmap calls to loop around and then crash.
A way to test this is by... -
-
-
-
-
- I believe this would be a very valuable feature (we discussed that ~1 year ago), particularly for testing environments.
Right now testing that a given API is Ractor-compatible is very hard, an API like the one described would make it ... -
- https://bugs.ruby-lang.org/issues/22276
-
- This is a messy one that originally surfaced as a flaky test in our Rails app, and required a lot of AI-assisted debugging to track down. This bug appears in the latest 4.0.7, though all of the original investigation was done against 3.3...
- 02:50 PM Revision 6e38174c (git): ZJIT: Specialize polymorphic sends with block literals (#18567)
- Reduce polymorphic sends on lobsters by ~850k. Unfortunately,
`send_block_arg_not_nil` goes up by ~500k. So the net benefit here is
~350k. The `send_block_arg_not_nil` needs investigating. - 02:39 PM Revision fb33630a (git): Get main Ractor to not take VM lock during local GC sweeping
- Sweeping unshareables generally doesn't modify global VM data structures. The exception
is rb_free_generic_ivar, which modifies the generic_fields_tbl. However, that table has
its own mutex protecting it. -
-
- Applied in changeset commit:git|0cdc4562e580bb104e03fe97553548bf20eca6e2.
----------
Fix assertion error when calling Array#initialize
[Bug #22333]
If we call Array#initialize again, it may trigger an assertion error since
the length ... -
- Thank you for the bug report, I have [a fix here](https://github.com/ruby/ruby/pull/18907).
-
- The following program:
```ruby
def assert_raise(*a) = (yield rescue nil)
MIN_SIZE = ENV.fetch('SMALL_ARRAY_MIN', 0).to_i
MAX_SIZE = ENV.fetch('SMALL_ARRAY_MAX', 16).to_i
ITERATIONS = ENV.fetch('SMALL_ARRAY_ITERATIONS', 100).to_i
... -
- [Bug #22333]
If we call Array#initialize again, it may trigger an assertion error since
the length is not reset before the capacity is changed. -
- Applied in changeset commit:git|85611dc266ff12ef7d311713f475ec1376f8ef86.
----------
[Bug #22324] APIs with observable side effects are not pure -
-
- Ruby and CPU heap doesn't make any sense when using the NoGC plan.
https://github.com/ruby/mmtk/commit/85f97bb0d9 -
- Making `RBIMPL_ASSERT_OR_ASSUME(needles_count > 0);` redundant.
-
- GCC was emitting a warning, because it couldn't understand that
`search->needles_count` was assumed not to be 0:
```
string.c:9673:36: warning: ‘matches’ may be used uninitialized [-Wmaybe-uninitialized]
9673 | const in... -
- @headius
You're right - my bad — the isolation there is incidental, not a guarantee. If the frame is captured as a shared proc, threads race on it badly.
But I think that actually restates my point rather than refuting it. There is no ... -
- > I'd gently push back on this, because switching to a thread already does exactly that today, and has for 20 years:
This is not actually true. Your example works that way because a thread clones its block's immediate frame. If the fram... -
- @headius
Just thinking out loud...
> Switching your code to a fiber shouldn't suddenly make them disappear, or should it?
I'd gently push back on this, because switching to a *thread* already does exactly that today, and has for... - 06:56 AM Revision 1c391eba (git): Do not flush the write buffer on close in sync mode. (#18902)
- When a stream is in sync mode (`sync == true`), writes go straight to the
operating system and Ruby's write buffer is not authoritative. Any bytes
left in the write buffer are therefore the result of writes made while
sync was disabled.
... -
06:18 AM Feature #22309: Allow super in a module method to work if method was called by refinement method super
- @shugo Thank you for the recommendation for implementing this correctly. I used the approach you recommended, and it ended up being much simpler than I expected, even if there were a couple tricky parts. One of those was how to store the...
-
- The following script causes a buffer overflow and returns corrupted strings
because it's reading past the end of the string buffer:
s = "x" + "l" * 3999
obj = Object.new
obj.define_singleton_method(:to_int) do
s.clear
... - 05:27 AM Revision 3b26b003 (git): Flush coalesced data in `io_binwritev` under sync mode. (#18905)
- `IO#write` with many arguments in sync mode was not observably atomic:
when the argument count exceeded `IOV_MAX`, the internal writev path
coalesced the trailing data into the write buffer and returned without
flushing, so nothing reach... -
05:00 AM Revision 21a472e8 (git): [ruby/rubygems] Always require the vendored SecureRandom in Gem::AtomicFileWriter
- `defined?(Gem::SecureRandom)` is already true while another thread is still loading the file and has not yet extended it with `Random::Formatter`, so a parallel installer worker could skip the require and fail with `NoMethodError` on `Ge...
-
- ruby/ruby#18574 makes $stdout and $stderr reassignment reach Kernel#puts and Kernel#warn inside a box. The box lanes run on a ruby-core master build, which has the fix.
https://bugs.ruby-lang.org/issues/21867
https://github.com/ruby/ru... -
- Fixes the following crash:
str = "hello" * 100
obj = Object.new
obj.define_singleton_method(:to_str) do
str.replace("")
"x"
end
str[/h.l/] = obj - 03:36 AM Revision 11f6d3d6 (git): Flush buffered data in `BasicSocket#close_write` before shutdown. (#18900)
- When a socket has buffered output (`sync == false`) and `#close_write` is
called, the write side was shut down via `shutdown(SHUT_WR)` without first
flushing the buffer. This had two bad consequences:
- the buffered bytes were silently ... - 03:15 AM Revision c55886f9 (git): Remove VM lock acquire during finalize_list and run_final
- Each objspace has its own finalizer list since RLGC landed so there aren't
concurrency issues here. This used to be a global table that all Ractors
shared. This affects only the default GC. -
- Applied in changeset commit:git|c914aeb622ce7337223cd5c23f30daf9f7d3e7c4.
----------
Fix use-after-free in String#encode when string modified
[Bug #22330]
The following script demonstrates a use-afer-free where we see corruption:
... -
- Thank you for the bug report, I have [a fix here](https://github.com/ruby/ruby/pull/18899).
-
- [Bug #22330]
The following script demonstrates a use-afer-free where we see corruption:
s = "あ" * 10_000
s.encode("US-ASCII", fallback: proc { |c| s.clear; "?" })
Raises may different errors such as:
"\xCB" followed by "\... -
- A relative Location died with `NoMethodError` because it was parsed on
its own instead of against the request URI, 308 fell through to "Bad
response", the non-https rejection printed the Location's credentials in
clear text, and an absol... -
- `net_http_args` calls `no_proxy?` whether or not a proxy is configured,
so a typo such as `https:/host` that URI parses with a nil host died with
`NoMethodError` instead of reaching the connection error that reports
the URI.
https://git... -
- `https://TOKEN:@host/` parses with an empty-string password, so `token?`
did not recognize it and `redacted` kept the token in clear text as
`https://TOKEN:REDACTED@host/`. `redact_credentials_from` also matched
the empty string and inse... - 02:36 AM Revision 5a8c2ac2 (git): [DOC] Update bundled gems list at 0772e42d0418cc4a138d8dfe7de936
- 02:35 AM Revision 0772e42d (git): Bump the github-actions group across 1 directory with 2 updates
- Bumps the github-actions group with 2 updates in the / directory: [ruby/setup-ruby](https://github.com/ruby/setup-ruby) and [taiki-e/install-action](https://github.com/taiki-e/install-action).
Updates `ruby/setup-ruby` from 1.322.0 to ... -
- Psych::Visitors::YAMLTree.create has existed since psych 2.0, so the shim was never defined on any supported Ruby.
https://github.com/ruby/rubygems/commit/242ca5848b
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> -
- Ruby 3.2 ships openssl 3.1, so PKey#public_to_der and SSLContext#min_version= are always available. get_public_key is documented to take a PKey, so the test that passed it a certificate now reads the key from the certificate directly.
h... -
- Symbol#name (3.0), Exception#detailed_message (3.2), Module#ruby2_keywords (2.7) and URI::Generic#hostname (1.9.3) exist on every supported Ruby, so the fallbacks were dead code.
https://github.com/ruby/rubygems/commit/ee203a9319
Co-Au... -
- GemVersionPromoter sorts every candidate newest first under --major, and --strict does not cap it at any major, so "next major version" never matched what it does.
https://github.com/ruby/rubygems/issues/8090
https://github.com/ruby/ru... -
-
- Gem::Specification#to_ruby interpolates the version into the `# stub:` comment without escaping, and #ensure_loadable_spec evals that output, so a line break in the version ends the comment and the rest runs as Ruby before the version it...
- 02:02 AM Revision e755b92a (git): Call dfree on non-thread-safe T_DATA with the VM barrier
- When running multiple Ractors, we can't call dfree of a non-thread-safe
T_DATA while other Ractors are running or doing GC work. If there are multiple
Ractors, we save the T_DATA object's `dfree` and `data` pointer in an entry. Once
the ... -
- Reopening, since this should stay open until removal occurs.
-
01:30 AM Bug #22273 (Closed): Aliasing doesn't interact well with Module#prepend
- Applied in changeset commit:git|1c13d07cfca130ae3c0631d24ef21680099c02a7.
----------
Emit deprecation warning for aliasing method in prepended module
Aliasing a method in a prepended module can result in a super
call going into a desce... -
-
- Aliasing a method in a prepended module can result in a super
call going into a descendant instead of an ancestor. Removal plan:
4.1: Deprecation warning
4.2: Warning even in non-verbose mode
4.3: Removal (target method lookup starts at... -
- Ruby will be emitting a deprecation warning for this shortly. Switch
to using super instead of an alias. -
- A browser that opens a connection and closes it without sending anything
left request_line nil, and the NoMethodError from splitting it took down
`gem signin` because the listener thread sets abort_on_exception. Close
such a connection a... -
- probe_pqc_handshake decides whether PQC works from the client side, but
its server thread only rescued OpenSSL errors. Anything else, a reset
included, was re-raised by the join in ensure, where the method's own
rescue no longer applies,... -
- The accept loop had no rescue at all, so a client that reset the
connection took it down with Errno::ECONNRESET and a client that hung up
without sending a request took it down with a NoMethodError on the nil
request line. Both then surf... -
- A client that dropped the connection mid handshake reached the accept
loop as Errno::ECONNRESET rather than OpenSSL::SSL::SSLError, so the
loop exited and stop_ssl_server re-raised it from join, turning teardown
into a test error. Window... -
- Under `RUBY_BOX=1`, `Kernel#pp` prints but leaves nothing behind in the box that called it.
```
$ RUBY_BOX=1 ruby --disable-gems -W:no-experimental -e 'pp({a: 1}); p $LOADED_FEATURES.grep(/pp\.rb/).size, defined?(PP)'
{a: 1}
0
nil... -
- MMTk's `make check` on macos-26 takes 27 minutes at the median and 37 at the 90th percentile, so the default 40 minute step limit sits inside the distribution of successful runs. Of the 61 MMTk macos-26 `make check` failures in the 1000 ...
-
- Under `RUBY_BOX=1`, loading a file into a box fails on `private`, `public`, `include`, `using`, `define_method` and `ruby2_keywords`.
```ruby
# toplevel.rb
private def hidden; end
puts "loaded"
```
```
$ RUBY_BOX=1 ruby -W:no-...
09/17/2026
-
- ioquatix (Samuel Williams) wrote in #note-24:
> @matz
> ...
It's a lovely idea but a very specific exception to introduce. These variables are specified to live in the nearest method frame, with no consideration for fiber locality. Lots ... -
- @matz
Fiber schedulers typically use `#transfer`, so preventing it would be problematic.
There is also an assumption being made that it's desirable for `svars` to leak out of enumerators, which may not always be the case, e.g.
... -
- When the string buffer is modified during the to_int call in Array#pack
with the r/R directive, it can cause an out-of-bounds write. For example,
the following script crashes:
str = "x" * 10000
obj = Object.new
obj.define_si... -
- Applied in changeset commit:git|29c206dd5522e13cdc7329721552dab6ac079842.
----------
[Bug #22280] Mark $? and $$ as box-ready
They are readonly virtual variables backed by per-thread/per-process
state, but user boxes cached the first r... -
- They are readonly virtual variables backed by per-thread/per-process
state, but user boxes cached the first read in the box gvar_tbl. The
cache stores a clone, and cloning Process::Status does not copy the
wrapped struct, so $? after Ker... -
- Assignment in a box only writes to box->gvar_tbl and leaves the global
entry with rb_gvar_undef_getter, so defined?($foo) returned nil while
$foo itself returned the assigned value. mkmf's have_devel? stops its
own recursion with `unless... -
- Applied in changeset commit:git|aca98ee9c7633ce0a1beffb605fc95c30837fb21.
----------
Mark stdio global variables as box-dynamic
$stdin, $stdout, $>, and $stderr live in ractor-local storage that
Kernel#puts, Kernel#warn, and friends re... -
- $stdin, $stdout, $>, and $stderr live in ractor-local storage that
Kernel#puts, Kernel#warn, and friends read directly, so assignments
cached in the Box gvar_tbl were invisible to them. Bypass the cache
like `$_` so the real setter updat... -
-
- The test helpers now boot the VM themselves via ensure_rubyvm(), so tests
no longer need a bare eval("nil") just to initialize ZJITState before
calling assert_compiles* or inspect. -
-
-
-
-
- Previously the Ractor courier rejected singleton class objects. However
we have existing tests of OpenStruct that those objects can be
Ractor-copied.
dup and clone, which were used prior to RLGC's merge, also just ignore
singleton class... -
- The previous commit integrated Marshal's hooks into the native courier
copying path. That should cover most cases.
One exception I've noticed that this doesn't implement is singleton
classes which have been extended by modules. I would ... -
- For some types (probably most notably Set), Marshal loading is defined
with compatibility functions. We should use those when set. -
- Previously the courier only carried the core types natively, and any
other object made the whole message fall back to Marshal.dump on the
sender and Marshal.load on the receiver. That serialized everything
reachable into bytes, so a Ract... -
- Previously any Regexp objects being sent would cause us to take the
Marshal path. -
- Previously rb_ractor_courier_materialize filled every non-hash node in
id order and then inserted hash entries in a separate pass in reverse
id order, on the assumption that ids are assigned depth-first so a key
always has a larger id th... -
- Hello, a heap UAF seems possible in string's encode function.
PoC
```
s = "あ" * 4000
s.encode("US-ASCII",
fallback: proc { |c|
s.replace("Z" * (... -
03:56 PM Feature #22329: Specify --disable-gem per Ruby Box
- The PR to implement this feature with options:
* have only `disable_gems` to disable all of 4 gems
* respect the process-wide selection and leave RubyGems an others are not loaded if `--disable=gems` and `disable_gems: false` are speci... -
- Currently user boxes loads gems listed in the gem_prelude.rb if those gems are enabled in the process-wide.
There is just only way to disable gems in boxes: specify `--disable-gems` in the entire process.
This proposes a new keyword ... - 03:51 PM Revision 99e1c066 (git): Fix Ractor._activated: don't use Module#prepend
- Use aliasing instead of Module#prepend with super to redefine `require`
when coming out of single ractor mode. There are bad interactions
between Module#prepend and aliasing and we have seen bugs related to
this in Rails CI: https://bugs... - 02:47 PM Revision 685aff92 (git): Update default gems list at cb7e0e62e255ce3622108bcd6934b3 [ci skip]
-
- https://github.com/ruby/io-console/commit/a44ab18ad9
-
- https://github.com/ruby/io-console/commit/668e5d85bb
- 02:12 PM Revision 6954f692 (git): ZJIT: Add validation pass for reducibility (#18754)
- Based on [ART's](https://android.googlesource.com/platform/art/+/refs/heads/main/compiler/optimizing/nodes.cc#999):
```c++
bool HLoopInformation::HasBackEdgeNotDominatedByHeader() const {
for (HBasicBlock* back_edge : GetBackEdges()) ... -
12:35 PM Bug #22273: Aliasing doesn't interact well with Module#prepend
- Besides the fundamental semantics of aliasing + prepending, I would like to stress that we have here yet another use case of Kernel#require decoration.
I don't want to hijack the ticket, but would like to leverage it to ask you: Don't... -
- string.c:9673:36: warning: ‘matches’ may be used uninitialized [-Wmaybe-uninitialized]
9673 | const int bitmap = _mm_movemask_epi8(matches[0]);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
str... -
- `update_default_gems` had no coverage, so nothing pinned the one thing it must not do to a gem checkout: swap it over to the ruby/ruby tree. The fixture drives it against local repositories, so the fetch it still needs stays offline.
Co... -
- `up` checked out ruby-core in every default gem repository to rebase it onto ruby-core/master and then checked the default branch back out, which walks the whole ruby/ruby tree twice per gem for a branch that nothing in this file reads. ...
-
- ruby/ruby owns this file along with ext/json/extconf.rb, which is already
excluded. Upstream json has no ext/json/ext/depend to copy back, so a
whole-tree sync deleted it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> -
08:28 AM Bug #22328 (Open): Should IO#wait_priority be deprecated and then removed?
- In #18655, IO#wait_readable, IO#wait_writable, IO#wait_priority, and IO#wait were moved from io/wait into core.
However, in #18655 note 11, Matz wrote:
> After the discussion, we will drop wait_priority, but others are accepted.
htt... -
- The call to rand (through RAND_UPTO) or to_int (through NUM2LONG) can run
aribtrary Ruby code, which can modify the source array. The size of the
array may shrink, which could cause an out-of-bounds read on the array.
We do check again t... - 07:00 AM Revision 22fda5cc (git): [DOC] Update bundled gems list at 91810913bd418c066ee77117642275
- 05:32 AM Revision 91810913 (git): Update default gems list at 9099b5f81a66db870ee56c1785913f [ci skip]
-
- https://github.com/ruby/resolv/commit/99464a354c
-
- A peer that goes away while another nameserver is being tried leaves a socket nobody is watching, so the loss only surfaces when the next request is written to it, as Errno::EPIPE. Config#resolv moves on to the next nameserver for Resolv...
-
- Three tests carry the same hand built reply, and the tests below need a
fourth.
https://github.com/ruby/resolv/commit/d01450fd36
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> -
05:10 AM Bug #22323: Incorrect nonnull contract changes rb_typeddata_inherited_p(NULL, NULL) behavior
- Oh I see the point. Thank you for the clarification and commit!
-
- Applied in changeset commit:git|b9d4fd7108b353fd58717c556757f0bce746c4cc.
----------
[Bug #22323] [DOC] Remove a stale internal note for typed data
Neither argument to `rb_typeddata_inherited_p` is intended to be NULL. -
- Neither argument is intended to be NULL. The `nonnull` annotation was added in 72627d85e337e5d8c7fa5738dc4ec7f253f0738e, which also annotated `RTYPEDDATA_TYPE` with `RBIMPL_ATTR_RETURNS_NONNULL()`. The expected `child` argument is the ...
-
- Thank you. By “the explicit equality behavior,” I meant this check at the
beginning of the inline implementation:
if (RB_LIKELY(child == parent))
return true;
The current header therefore gives three conflicting signa... -
- Thank you for the report.
yqtian (Yongqiang Tian) wrote:
> Could you confirm whether `(NULL, NULL)` is an intended input for this
> ...
I don't think it is intended.
What do you mean by "the explicit equality behavior"? -
- Label.split maps both "." and "" to the root label list, so a resolv.conf
carrying `search .` produces a search entry that reproduces the absolute name.
Tidying up ndots handling dropped the include?(fname) guard that used to
suppress th... -
- The current code works, but the intention was not clear at first glance.
https://github.com/ruby/resolv/commit/deaf5ab649 -
- When search list is not explicitly specified, Resolv::DNS checks the
local domain name of the running system. If it is not set either, the
root domain ['.'] is used.
This is conceptually correct, but the code handles the absolute name
s... -
- Add more assertions to actually test that the ndots option is working
as expected: the option controls the threshold that the search list
should be attempted earlier or later.
https://github.com/ruby/resolv/commit/a6b5856ef4 -
- I've added a verbose-mode warning: https://github.com/ruby/ruby/pull/18740
-
- Neither argument to `rb_typeddata_inherited_p` is intended to be NULL.
-
-
- `alias` and `Module#alias_method` in a module fall back to searching
the method from Object when the module and its ancestors do not have
it. The fallback assumes Object is the root class, which is no longer
true since BasicObject was in... - 03:31 AM Revision 6dabc1cf (git): Update default gems list at 028dcaa3c9513aa61903989b979e41 [ci skip]
-
- https://github.com/ruby/net-protocol/commit/c1f66145f7
-
- LOG_off had no counterpart on the exception path, so a raise from the
block or from the source silenced the connection's debug_output for
good. The two success cases go with them because the restructure
rewrote the method that had no cov... - 03:25 AM Revision 550591bb (git): [ruby/net-protocol] Restore message logging and clear write state on exceptional exits
- https://github.com/ruby/net-protocol/commit/a1bd220aad
-
- The case had no branch for it, so the retry loop spun without waiting
and without ever reaching the write_timeout. The fake refuses the write
until it has been waited on, which turns a regression into a failure
rather than a hang.
https... - 03:25 AM Revision d9c88e84 (git): [ruby/net-protocol] Wait for readability when an SSL write requests it
- https://github.com/ruby/net-protocol/commit/85361b97cd
-
- Before the check, read(-1) walked @rbuf_offset backwards and a later
read_all handed back bytes an earlier read had already consumed.
https://github.com/ruby/net-protocol/commit/d89c7b148b
Co-Authored-By: Claude Opus 5 <noreply@anthrop... - 03:25 AM Revision 48c7a6ae (git): [ruby/net-protocol] Reject negative read lengths before changing the buffer
- https://github.com/ruby/net-protocol/commit/44b2dcb2f4
- 02:33 AM Revision e6ec1555 (git): [DOC] Update bundled gems list at b06e76429cd6a79505bbc647e2d342
- 02:31 AM Revision b06e7642 (git): Bump ruby/setup-ruby in the github-actions group across 1 directory
- Bumps the github-actions group with 1 update in the / directory: [ruby/setup-ruby](https://github.com/ruby/setup-ruby).
Updates `ruby/setup-ruby` from 1.321.0 to 1.322.0
- [Release notes](https://github.com/ruby/setup-ruby/releases)
- ... -
- The call to rb_to_int may run aribtrary Ruby code and thus cause String#setbyte
to read past the end of the string. For example, the following script crashes:
s = "a" * 10_000_000
obj = Object.new
obj.define_singleton_method... -
- ruby_3_3 commit:52c54450f9acf0762089e21f0338dc393ec547ef merged revision(s) commit:190b017fc6c21ff7b61c2b5ece0294785e4a4ca2.
-
- ruby_3_3 commit:0c8ab2a111722f38b58738dd3d8fb77b79c11249, ruby_3_4 commit:5a42e10000e344f2143b772b18b700a9f8e3c9cf and ruby_4_0 commit:d50f404ca12c04e8bc3e8dd805879a660258b5c4 merged revision(s) commit:6bfde8d2a239a9e8549de1ae29acecdefdf...
-
- To account for cases where binaries for older OS versions are built
using the new SDK, check at runtime whether the API functions are
available actually. -
- [Bug #21703]
RUBY_CRASH_REPORT does not work in some cases when shelling out on Linux.
For example, given the following shell script dump.sh:
#!/usr/bin/env bash
cat > /tmp/crash
And we see it fails like this:
$ RUBY_CRAS... -
- Instead of using `assert_in_out_err` directly.
-
-
-
- OK. I submitted https://github.com/ruby/ruby/pull/18848 to fix this in master (same as https://github.com/ruby/ruby/pull/17633 but squashed and with fixed specs). It leaves Ruby 4.0 behavior unspecified. After it is merged, I'll submit ...
- 12:51 AM Revision 548b5abc (git): Haiku: use ucontext.h only on x86/x86-64
- This follows the other operating systems. Without it, it won't build on arm64 and riscv64 ports.
-
12:32 AM Bug #22327 (Open): RubyVM::ZJIT.reset_stats! aborts the process (Rust panic) when ZJIT is not enabled
- `RubyVM::ZJIT.reset_stats!` aborts the whole process with a Rust panic when it is called while ZJIT is not enabled. The other stats methods (`stats`, `stats_string`, `stats_enabled?`) just return `nil`/`false` in the same situation, and ...
-
- Array#values_at can execute arbitrary code during execution which may
modify the source array. This can cause a buffer overflow if the new length
of the array has decreased. The following script crashes:
a = (0..100_000).to_a
ob...