Activity
From 06/28/2024 to 07/04/2024
07/04/2024
-
07:35 PM Feature #20405: Inline comments
- This could be particularly useful for projects like `rbs-inline`.
-
05:40 PM Bug #18061: Execshield test: libruby.so.N.N.N: FAIL: property-note test because no .note.gnu.property section found
- fweimer (Florian Weimer) wrote in #note-48:
> The change in https://github.com/ruby/ruby/pull/11081/files does not look correct to me because it enables shadow stack and indirect branch tracking in the ELF markup. But the changes to `Co... -
07:28 AM Bug #18061: Execshield test: libruby.so.N.N.N: FAIL: property-note test because no .note.gnu.property section found
- The change in https://github.com/ruby/ruby/pull/11081/files does not look correct to me because it enables shadow stack and indirect branch tracking in the ELF markup. But the changes to `Context.S` merely enable indirect branch tracking...
-
- ioquatix (Samuel Williams) wrote in #note-46:
> BTW, IIUC, if we can use CET instead of the shadow stack, it's a far simpler and more efficient solution, so I'm strongly in favour of that direction (assuming it's viable).
If I understan... -
01:25 PM Revision c480dec1 (git): Fix RUBY_FREE_AT_EXIT with ASAN
- When Ruby is built with ASAN and RUBY_FREE_AT_EXIT is enabled, the
following error occurs:
READ of size 8 at 0x74c666610020 thread T0
#0 0x593b6712ecc6 in RB_BUILTIN_TYPE include/ruby/internal/value_type.h:191:30
#1 ... -
11:00 AM Misc #20574: DevMeeting-2024-07-11
- * [Bug #20606] `Thread#thread_variable_get` and `Thread#thread_variable?` don't raise `TypeError` exception for incorrect argument value when thread-local variable storage is not initialised yet
* These methods raise `is not a symbol ... -
10:02 AM Bug #20606: Thread#thread_variable_get, Thread#thread_variable? and Thread#[] methods handle non-String/Symbol parameter values differently
- > I'm not opposed to changing the behavior, but raising an exception when one was not previously raised presents backwards compatibility issues. We would probably have to warn in 3.4 and not raise until 3.5 (or later).
I don't think a... -
- @jeremyevans0 The specs were added in https://github.com/ruby/spec/pull/1127 (so not written by Andrii, but reviewed by him).
Adding specs is how we discover this kind of strange behavior.
But it's not always clear if intentional or no... -
03:33 AM Revision 478ada0e (git): [rubygems/rubygems] Handle OpenSSL::SSL::SSLError in local https server
- https://github.com/rubygems/rubygems/commit/ccbbe84d77
-
- https://github.com/rubygems/rubygems/commit/c31a1592ee
-
- https://github.com/rubygems/rubygems/commit/c538558522
-
- https://github.com/rubygems/rubygems/commit/29f71ee6d2
-
- https://github.com/rubygems/rubygems/commit/c755bcde62
-
- https://github.com/rubygems/rubygems/commit/ee45ebb5c3
-
- https://github.com/rubygems/rubygems/commit/3ae4358024
-
- https://github.com/rubygems/rubygems/commit/21319eccac
-
- https://github.com/rubygems/rubygems/commit/69bd962b56
-
- https://github.com/rubygems/rubygems/commit/43f98b787b
-
- https://github.com/rubygems/rubygems/commit/52db9fd501
-
- https://github.com/rubygems/rubygems/commit/d3ee8d2d3b
-
03:11 AM Revision 0fe024d0 (git): Fix the return value of `Integer#downto` called with a block
- As the document states, it should return `self`, not `nil`.
Fix up of f4b313f7338f5fbe37f73aae29f70aeb474f7f5b. - 02:55 AM Revision 70be2f47 (git): Bump ruby/setup-ruby from 1.183.0 to 1.184.0
- Bumps [ruby/setup-ruby](https://github.com/ruby/setup-ruby) from 1.183.0 to 1.184.0.
- [Release notes](https://github.com/ruby/setup-ruby/releases)
- [Changelog](https://github.com/ruby/setup-ruby/blob/master/release.rb)
- [Commits](http...
07/03/2024
-
11:40 PM Bug #18061: Execshield test: libruby.so.N.N.N: FAIL: property-note test because no .note.gnu.property section found
- BTW, IIUC, if we can use CET instead of the shadow stack, it's a far simpler and more efficient solution, so I'm strongly in favour of that direction (assuming it's viable).
-
- Thanks, I agree with your proposed change.
-
- 11:37 PM Revision d9398ac4 (git): Use `__CET__` macro to enable IBT support. (#11081)
- According to the GCC documentation [1], the macro `__CET__` is dfined
when `-fcf-protection` compiler option is used. Therefore use this macro
to enable IBT support instead of special casing for `__OpenBSD__`.
[1]: https://gcc.gnu.org/o... -
11:10 PM Revision b160a78d (git): YJIT: Remove done TODO, fix indent
- Type check now done in rb_iseqw_to_iseq().
-
11:01 PM Bug #20597: `eval('break if false')` should raise SyntaxError but retuns nil
- I submitted a pull request to fix this (but prism still needs a related fix) https://github.com/ruby/ruby/pull/11099
-
- This change would be simple to implement, and I agree the fiber/thread variable inconsistency seems odd. There are specs for the current behavior. However, I checked and the specs appear to be added by you only a couple days ago. It se...
-
- The `Thread#thread_variable_get`, `Thread#thread_variable?` and `Thread#[]` methods handle the `key` parameter that is not a String or a Symbol in different way but I would expect them to be consistent and raise an exception.
When no ... -
- I submitted a pull request to make the error message consistent: https://github.com/ruby/ruby/pull/11097
-
- `Thread`'s methods to manipulate thread-local variables raise `TypeError` exceptions with slightly different messages when the `key` argument is not a String or a Symbol. The message is either `is not a symbol` or `is not a symbol nor a ...
-
08:00 PM Revision f5dfadf3 (git): [DOC] Doc for Integer#floor (#11077)
-
07:32 PM Revision b974c846 (git): Move Array#map to Ruby
- Improves activerecord by about 1% on the interpreter:
```
before: ruby 3.4.0dev (2024-07-03T18:40:10Z master f88841b8f3) [arm64-darwin23]
after: ruby 3.4.0dev (2024-07-03T18:41:14Z ruby-map 6c0df4eb32) [arm64-darwin23]
------------ --... -
-
-
-
- This speeds up the mail benchmark by about 7% on the interpreter:
```
before: ruby 3.4.0dev (2024-07-03T17:01:41Z master f4b313f733) [arm64-darwin23]
after: ruby 3.4.0dev (2024-07-03T17:45:50Z ruby-select de282cacd5) [arm64-darwin23]
-... -
-
01:42 PM Bug #20605 (Closed): PRs with backported fix for https://github.com/ruby/ruby/pull/11036 to Ruby 3.3 and 3.2
- Hey! I work for Datadog on the Ruby profiler part of the [datadog (previously ddtrace)](https://github.com/datadog/dd-trace-rb) gem.
Recently we merged a very small tweak to master to make ensure compilers did the right thing always: ... -
-
-
-
05:17 PM Revision 7fe5f0a1 (git): [ruby/irb] Introduce cd command
- (https://github.com/ruby/irb/pull/971)
It's essentially a combination of pushws and popws commands that are
easier to use.
Help message:
```
Usage: cd ([target]|..)
IRB uses a stack of workspaces to keep track of context(s), with `pu... -
- Speeds up ChunkyPNG.
The interpreter is about 70% faster:
```
before: ruby 3.4.0dev (2024-07-03T15:16:17Z master 786cf9db48) [arm64-darwin23]
after: ruby 3.4.0dev (2024-07-03T15:32:25Z ruby-downto 0b8b744ce2) [arm64-darwin23]
--------... -
- **What does this PR do?**
This PR tweaks the `vm_push_frame` function to add an explicit compiler
fence (`atomic_signal_fence`) to ensure profilers that use signals
to interrupt applications (stackprof, vernier, pf2, Datadog profiler)
c... -
-
- kjtsanaktsidis (KJ Tsanaktsidis) wrote in #note-4:
> and in fact I would go a step furthre and delete `$ASFLAGS` in `configure.ac` and `Makefile.in` entirely. It's late now but I'll try and put up a PR for this tomorrow or Friday.
I ... -
11:58 AM Bug #20601: Configuration flags are not properly propagated to assembler
- Just one more thing: I was worried for a moment that trying to assemble with `$CFLAGS` would cause problems if you had some C-language specific things in cflags, but actually it seems to work fine - even if you have something like `-std=...
-
- OK, I poked at this a bit tonight. I have quite a few thoughts but they're kind of disorganised.
The short version of this is: I think you're right; we should pass the full set of `$CFLAGS $CPPFLAGS` when calling `$(CC)` on a .S file, a... -
- Is something like that really needed? From my POV, the issue is that originally, there were `.s` files processed by Assembler, therefore the `$(ASFLAGS)` were needed. But now we have `.S` files pre-processed by compiler:
https://githu... -
- I think we should do what we did for the pac-ret stuff, that Florian recommended, right? Run tests against CFLAGS in configure.ac, and then define special macros?
https://github.com/ruby/ruby/blob/93b19d56de64fdee790a96ddf96fcd08d889a... -
- Applied in changeset commit:git|51bd816517941798c63e587a5a9f3caf69cd510e.
----------
[Feature #20470] Split GC into gc_impl.c
This commit splits gc.c into two files:
- gc.c now only contains code not specific to Ruby GC. This includes... -
03:16 PM Revision 786cf9db (git): array.c: Remove outdated assertions
- Following [Feature #20589] it can happen that we change the
capacity of a frozen array, so these assertions no longer make
sense.
Normally we don't hit them because `Array#freeze` shrinks the
array, but if somehow the Array was frozen u... -
- Closing given https://github.com/ruby/ruby/pull/11030 was merged yesterday.
However it caused some assertions failures in corner cases, so I opened a followup: https://github.com/ruby/ruby/pull/11092 -
-
-
-
- This commit implements support to load Ruby's current GC as a DSO.
-
- This commit splits gc.c into two files:
- gc.c now only contains code not specific to Ruby GC. This includes
code to mark objects (which the GC implementation may choose not to
use) and wrappers for internal APIs that the implementa... -
- https://github.com/ruby/shellwords/commit/53e6ef2e71
-
- https://github.com/ruby/shellwords/commit/fa50d40a33
-
- https://github.com/ruby/shellwords/commit/08c87a964e
-
- **What does this PR do?**
This PR tweaks the `vm_push_frame` function to add an explicit compiler
fence (`atomic_signal_fence`) to ensure profilers that use signals
to interrupt applications (stackprof, vernier, pf2, Datadog profiler)
c... -
08:54 AM Revision 4d4ac001 (git): [ruby/openssl] Add SSLSocket#readbyte
- Companion to getbyte but raise EOFError
Similar to https://github.com/ruby/openssl/pull/438
https://github.com/ruby/openssl/commit/c40f70711a -
- It doesn't look like there was a test added for this bug, so I'm adding
it.
Code is from here:
https://web.archive.org/web/20160908192307/http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-dev/24445 -
07/02/2024
-
- * Speed up chunkypng benchmark
Since d037c5196a14c03e72746ccdf0437b5dd4f80a69 we're seeing a slowdown
in ChunkyPNG benchmarks in YJIT bench. This patch addresses the
slowdown. Making the thread volatile speeds up the benchmark by 2 or 3... -
09:53 PM Bug #20603: Rails Active Job test gets segmentation fault
- Confirmed b2b8306b46 addresses this issue. Thanks for the fix.
```
$ ruby -v
ruby 3.4.0dev (2024-07-02T18:48:43Z master b2b8306b46) [x86_64-linux]
$ AJ_ADAPTER=test bin/test "test/cases/test_helper_test.rb" -n "/^(?:PerformedJobsTe... -
05:37 PM Bug #20603: Rails Active Job test gets segmentation fault
- I opened a fix https://github.com/ruby/ruby/pull/11086
-
- Rails Active Job test sometimes gets segmentation fault.
https://buildkite.com/rails/rails-nightly/builds/723#01906af3-7abc-4f9a-bfd7-c3ff49bbf6cf/1363-1370
This segmentation fault does not always reproduce so I have not found which ... - 07:55 PM Revision 3407565d (git): YJIT: Use a special breakpoint address if one isn't explicitly supplied in order to support natural line stepping. (#11083)
- Use a special breakpoint address if one isn't explicitly supplied in order to support natural line stepping.
ARM64 will not increment the program counter (PC) upon hitting a breakpoint instruction. Consequently, stepping through code wi... -
07:54 PM Revision 2830a6ae (git): [rubygems/rubygems] Only allow valid values for --test, --ci, and --linter options
- https://github.com/rubygems/rubygems/commit/d4360c9032
-
07:53 PM Revision cd57c129 (git): [rubygems/rubygems] Only remove current platform when added right before resolution
- If it has been explicitly added by `bundle lock --add-platform`, we
don't want to remove it.
https://github.com/rubygems/rubygems/commit/09cf921dd6 -
- https://github.com/rubygems/rubygems/commit/18b38ca6b4
-
- https://github.com/rubygems/rubygems/commit/1c80c6072f
-
- https://github.com/rubygems/rubygems/commit/0236cb6191
-
- Always treat forwarding as a complex call.
-
- When we forward calls to C functions if the callsite is a forwarding
site it might not always be a splat, so we can't use the fast path.
Fixes:
[ruby-core:118418] - 06:34 PM Revision 4d94d28a (git): YJIT: Inline simple ISEQs with unused keyword parameters
- This commit expands inlining for simple ISeqs to accept
callees that have unused keyword parameters and callers
that specify unused keywords. The following shows 2 new
callsites that will be inlined:
```ruby
def let(a, checked: true) = ... -
- While working on a separate issue we found that in some cases
`ary_heap_realloc` was being called on frozen arrays. To fix this, this
change does the following:
1) Updates `rb_ary_freeze` to assert the type is an array, return if
alread... -
05:18 PM Bug #20604 (Rejected): Performance regression in C++ extensions due to lack of optimization flags by default since Ruby 2.7
- I found a significant performance regression in `sassc` gem when comparing Ruby 2.6 and later, that the extension is running more than 10x slower.
I have tracked it down to this commit: https://github.com/ruby/ruby/commit/733aa2f8b578... -
- I have opened https://bugs.ruby-lang.org/issues/20588 and the test case has been fixed via 4cbc41d5e5cb6793174d5964975fdb4470323ca1 , It looks like there are another test case that still reproduces against the latest master branch like c...
-
- This might be the medicine:
~~~patch
$ git diff
diff --git a/template/Makefile.in b/template/Makefile.in
index 2d232d7925..31e5ae5edb 100644
--- a/template/Makefile.in
+++ b/template/Makefile.in
@@ -449,7 +449,7 @@ $(srcdir)/enc... -
- Looking into #18061, one of the issues is that the `.S` files are not processed with the correct flags. For example to have the CET enabled, the `-fcf-protection` should be used to preprocess the `coroutine/amd64/Context.S`.
First I t... -
-
- As a small note on this, it's typically better to check `$VERBOSE` level and if the category is enabled before even generating/formatting the message String, for performance reasons.
So if that's always done the check in Warning.warn wou... -
- (https://github.com/ruby/rdoc/pull/1114)
It's hard to distinguish code object classes by their file names alone.
And given that we have 18 such classes, it'd make the codebase a lot
easier to understand if we grouped them into a single ... -
- `test/fiddle/test_function.rb`
(https://github.com/ruby/fiddle/pull/141)
https://github.com/ruby/ruby/pull/11048#issuecomment-2185630854
---------
https://github.com/ruby/fiddle/commit/a245f19763
Co-authored-by: Yusuke Endoh <mame@ru... -
06:00 AM Bug #20592: Interrupting Addrinfo causes Segmentation fault on alpine
- @nagachika @k0kubun When you backport this patch, please backport commit:d8c6e91748871ab2287d7703347847fe18a292d2 as well. FYI.
-
04:47 AM Misc #20600 (Closed): @ruby-lang.org has been migrated to Cloudflare
- @ruby-lang.org has been migrated to Cloudflare, because ruby-lang.org was migrated from Google Domains to Squarespace, and Squarespace's e-mail forwarding doesn't work with custom MX records including subdomains:(
I and Shibata-san ar... -
07/01/2024
-
- This issue does not reproduce since a2c27bae9653a817b4e5f699f421836f8f97410b .
However, similar issue like #20588 happened again. https://buildkite.com/rails/rails-nightly/builds/727#01907019-eb31-495b-9b48-53598cac9928/1293-1302 I'll f... -
- vo.x (Vit Ondruch) wrote in #note-41:
> Given that [GCC documentation](https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html#index-fcf-protection) mentions that specifying the `-fcf-protection` compiler option defines the __CE... -
- And FTR, [this](https://github.com/gcc-mirror/gcc/blob/8088374a868aacab4dff208ec3e3fde790a1d9a3/gcc/config/i386/cet.h) is the `cet.h` content
-
- As can be seen, I am testing this with Ruby 3.3.2. The upstream situation is, however, already a bit different since [#10987](https://github.com/ruby/ruby/pull/10987) (I actually took the inspiration from that PR). Given that [GCC docume...
-
- Forgot to mention that Ruby needs to be built with `--disable-yjit`, because Rust is another problem:
https://bugzilla.redhat.com/show_bug.cgi?id=2284605#c1 -
- So I am experimenting with this simple patch:
~~~patch
diff --git a/coroutine/amd64/Context.S b/coroutine/amd64/Context.S
index fcc4b67bdc..b364fc06a1 100644
--- a/coroutine/amd64/Context.S
+++ b/coroutine/amd64/Context.S
@@ -4,6... -
05:13 PM Bug #20099: Ruby 3.3.0 segfaults on s390x musl libc (Alpine Linux) when built with -O3 (default optflags)
- I've tried to reproduce the bug on Linux One s390x machine(using docker image alpinelinux/alpine-gitlab-ci with ruby 3.3.0 and 3.3.3) and tests succeeded without fault.
The build with -O3 succeed also(https://gitlab.alpinelinux.org/hau... -
03:22 PM Revision 0ac9c6dc (git): [ruby/prism] Fix up find and array patterns with multiple rests
- https://github.com/ruby/prism/commit/035890a839
-
- https://github.com/ruby/prism/commit/bedc4585ed
-
-
01:11 AM Bug #20596: [BUG] unexpected rb_parser_ary_data_type (0) for script lines
- Applied the patch to master branch https://github.com/ruby/ruby/pull/11076.
-
03:11 AM Revision 30a8dbc8 (git): maximum showing line number on btest failure
- Some tests in btest uses long src for btest and it is harmful to
check the results. This patch introducing the limitation how many
lines of code is shown on failure. - 02:53 AM Revision 89466f4e (git): Bump ruby/setup-ruby from 1.180.1 to 1.183.0
- Bumps [ruby/setup-ruby](https://github.com/ruby/setup-ruby) from 1.180.1 to 1.183.0.
- [Release notes](https://github.com/ruby/setup-ruby/releases)
- [Changelog](https://github.com/ruby/setup-ruby/blob/master/release.rb)
- [Commits](http... - 02:52 AM Revision 1289782d (git): Bump github/codeql-action from 3.25.10 to 3.25.11
- Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.25.10 to 3.25.11.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.... - 02:39 AM Revision c46af61b (git): [ruby/irb] fix typos in the `Index of Command-Line Options`
- https://github.com/ruby/irb/commit/3512020f1c
-
02:08 AM Revision 15ceb95d (git): [ruby/rdoc] Fix some typos (https://github.com/ruby/rdoc/pull/1129)
- * constist ==> consist
* Tidyness ==> Tidiness
* Currentry ==> Currently
* valus ==> values
https://github.com/ruby/rdoc/commit/8412705721 -
06/30/2024
-
05:13 PM Revision de2d9c8e (git): [ruby/irb] Allow assigning and using local variable name conflicting
- with command
(https://github.com/ruby/irb/pull/961)
https://github.com/ruby/irb/commit/00603d470f - 06:59 AM Revision c4baf3b3 (git): Update bundled gems list as of 2024-06-29
-
- I think this change fixes the following assertion failure:
```
[BUG] unexpected rb_parser_ary_data_type (2114076960) for script lines
```
It seems that `ast_value` is collected then `rb_parser_build_script_lines_from`
touches invalid m...
06/29/2024
-
- Similarly to splat arrays, we shouldn't expand splat kwargs.
[ruby-core:118401] -
03:48 PM Bug #20596: [BUG] unexpected rb_parser_ary_data_type (0) for script lines
- Just saw the same in Puma CI: https://github.com/puma/puma/actions/runs/9725211941/job/26842252268?pr=3418#step:4:115
-
06/28/2024
-
- 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... -
- 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.
-
- 02:11 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... -