Activity
From 06/02/2021 to 06/08/2021
06/08/2021
-
10:37 PM Bug #16782 (Closed): `lock': deadlock; recursive locking (ThreadError) in 2.7.1
- I have not seen this issue for a long time so it must be fixed.
-
- I agree we should introduce an abstract rich model for this.
https://bugs.ruby-lang.org/issues/13383 is a similar problem.
It makes sense to me to introduce some kind of `Backtrace::Location#source_location`.
We should stop thin... -
07:40 PM Feature #17930: Add column information into error backtrace
- mame (Yusuke Endoh) wrote in #note-5:
> In fact there is no enough information because `Thread::Backtrace::Location` only knows path and node_id.
This seems an implementation detail, not something that should decide what the API is.
... -
12:18 PM Feature #17930: Add column information into error backtrace
- Ahh, I've tried but I found that the current prototype draws an underline from "the end column of the receiver of the call" to "the end column of the call", so four integers are clearly not enough. Hardcoding such a location looks very a...
-
- I've created a PR for did_you_mean: https://github.com/ruby/did_you_mean/pull/151
Eregon (Benoit Daloze) wrote in #note-4:
> There is no point to use `RubyVM::AbstractSyntaxTree` when all the information is already in the `Thread::Ba... -
- > I talked with ko1 (Koichi Sasada), and we agreed that allowing RubyVM::AbstractSyntaxTree.of(Thread::Backtrace::Location) is a good approach to allow this feature.
I strongly disagree, RubyVM is a CRuby-specific API.
This effective... -
- Thank you @matz .
I talked with @ko1, and we agreed that allowing `RubyVM::AbstractSyntaxTree.of(Thread::Backtrace::Location)` is a good approach to allow this feature.
```ruby
def target
RubyVM::AbstractSyntaxTree.of(caller_lo... -
06:31 PM Revision 9f110ced (git): Add multi-heap support to gc_marks_wb_unprotected_objects
-
05:20 PM Bug #17767 (Closed): `Cloned ENV` inconsistently returns `ENV` or `self`
- Applied in changeset commit:git|117310bdc00236c0a7676616ce25b5106775dabc.
----------
Make ENV.clone warn and ENV.dup raise
ENV.dup returned a plain Object, since all of ENV's behavior is
defined in ENV's singleton class. So using dup ... - 05:19 PM Revision f9e13cda (git): * 2021-06-09 [ci skip]
-
- ENV.dup returned a plain Object, since all of ENV's behavior is
defined in ENV's singleton class. So using dup makes no sense.
ENV.clone works and is used in some gems, but it doesn't do what
the user expects, since modifying ENV.clone...
06/07/2021
-
- https://github.com/ruby/psych/commit/542cf9754f
-
09:54 PM Bug #17941: POSIX timers are not fork-safe on 2.7.x
- Here is a small C program which aborts on Ruby 2.7+ on Linux but works on Ruby 2.5:
```
#include <ruby.h>
#include <stdio.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <unistd.h>
#include <signal.h>
int main(int argc, char ... -
- I have an application which runs under a pre-fork-model web server (uwsgi).
When a worker exits, SIGCHLD will be delivered to the parent process, which is handled by `sighandler` and `rb_thread_wakeup_timer_thread`. This calls `ubf_ti... -
08:14 PM Bug #17940: Bus Error
- There is too little information to judge now, but my prejudication is that this is the same issue as in #17922 (and a whole bunch of other similar issues for a specific extension.)
-
07:21 PM Bug #17940 (Feedback): Bus Error
- In general, if you are using FFI and get a Bus Error, it is unlikely to be a bug in Ruby. It is more likely this is an issue where third party software was written for x86 and doesn't work correctly on M1 Macs. It's hard to determine ex...
-
07:01 PM Bug #17940 (Closed): Bus Error
- /Library/Ruby/Gems/2.6.0/gems/ffi-1.15.1/lib/ffi/library.rb:275: [BUG] Bus Error at 0x00000001050bc000
ruby 2.6.3p62 (2019-04-16 revision 67580) [universal.arm64e-darwin20]
-- Crash Report log information ----------------------------... - 06:47 PM Revision a49a08ab (git): * 2021-06-08 [ci skip]
-
06:47 PM Revision 459a9f82 (git): Don't create objects during GC
- If we crash during GC, allocating new objects in the segv handler can
cause an infinite loop. This commit is to avoid creating new objects in
the crash handler -
- At the May 2021 developer meeting, it was decided that `ENV.dup` should raise and `ENV.clone` should warn, so I closed my previous pull request and added a pull request for that behavior: https://github.com/ruby/ruby/pull/4557
-
04:51 PM Feature #17938: Keyword alternative for boolean positional arguments
- > This applies for any argument really. No other way than to read the docs (or know from a previous reading) to answer that precisely.
I get what you're saying, but I don't think it's a fair comparison. In the following example I think ... -
01:46 PM Feature #17938: Keyword alternative for boolean positional arguments
- We need to name keyword arguments consistently among similar methods. Otherwise, we will face confusion as you mentioned, I agree.
Matz.
-
- Another problem is
```ruby
object.respond_to?(:symbol, include_all: false)
object.methods(regular: true)
```
does not solve anything.
It's unclear what `include_all` or `regular` mean without looking at the docs.
And if one looks ... -
- Although it requires migration complexity and ugliness for the time being, I think it's OK.
We may need to investigate the performance of method calls with optional keyword arguments since some methods (e.g. `respond_to?`) are called o... -
- * [Bug #17561] The timeout option for Addrinfo.getaddrinfo is not reliable on Ruby 2.7.2 (jeremyevans0)
* Usage of getaddrinfo_a was reverted before 3.0, but still causes problems in 2.7.
* Do we want to revert commit 0e9d56f5e73ed... -
- * [Feature #12913] A way to configure the default maximum width of pp (mame)
* How about `pp(obj, pp_out: $stderr, pp_width: 100)`? -
04:04 PM Misc #17936: Run Travis at push on forked repository
- I sent the PR here.
https://github.com/ruby/ruby/pull/4556
-
- This was discussed at the May 2021 developer meeting. @naruse is against breaking backwards compatibility for this in 3.1, so I'll repropose this after the release of 3.1.
-
- The attached patch is essentially a cleanup of that by @Biswa96 from GitHub. With this `./configure rb_cv_msvcrt=ucrt140` should work for Mingw-w64 with the default crt being UCRT [1].
[1] https://sourceforge.net/p/mingw-w64/mailman/m... -
- I agree with the idea too. Let's see how it works.
Matz.
-
11:44 AM Revision 26fcec5a (git): [DOC] fix metacharacter markups [ci skip]
-
-
-
10:24 AM Revision bb4a1020 (git): Do not use YAML module in tests of Psych
-
-
- https://github.com/ruby/psych/commit/4049939006
-
10:15 AM Revision fd6225c7 (git): [ruby/psych] Implement YAML.safe_dump to make safe_load more usable.
- In case where Psych is used as a two way serializers,
e.g. to serialize some cache or config, it is preferable
to have the same restrictions on both load and dump.
Otherwise you might dump and persist some objects payloads
that you late... -
- YAML.load and YAML.safe_load are different a little; the former allows
Symbol by default but the latter doesn't. So YAML.load_file and
YAML.safe_load_file should reflect the difference.
Fixes #490
https://github.com/ruby/psych/commit/f... -
10:09 AM Feature #17853: Add Thread#thread_id
- Thank you, @naruse and @nobu!
-
- Some certificates on Unicode.org have expired.
-
- Unfortunately it often fails depending on the status of www.unicode.org.
https://github.com/mame/ruby/runs/2761346164?check_suite_focus=true
```
./tool/downloader.rb:248:in `rescue in download': failed to download 12.1.0/ucd/UnicodeData... -
- So, some people (including @ko1) have a custom to pass keyword arguments to Kernel#pp:
```
foo = 42
bar = 43
pp(foo: foo, bar: bar) #=> {:foo=>42, :bar=>43}
```
Unfortunately, introducing keyword arguments to `Kernel#pp` breaks... - 01:23 AM Revision a9d76172 (git): * 2021-06-07 [ci skip]
- 01:22 AM Revision ad22e889 (git): Fix Typo
-
- This segfault issue looks like #17518, which has been already fixed in 3.0.1. I have no idea about the other issue (puma hungs forever).
06/06/2021
-
07:48 PM Bug #17939 (Rejected): Segmentation fault at Ruby 3.0.0 w/ CGI.escapeHTML
- It reproduces with 2.7, and it seems like the patched gem allocates a lot of memory. Because I'm still in the middle of making the change to this gem, I'll re-report this after finishing up the patch.
-
- I tried to check if this reproduces with 3.0.1 as well, but it seems to hit another issue that puma hangs forever and I wasn't able to confirm it.
-
- When I run https://github.com/k0kubun/gitstar-ranking with a Gemfile modified to use haml.gem at https://github.com/k0kubun/haml/commit/e3d921e5f6287aa4a4aec15f096a78dd56bd8f46 instead of hamlit.gem, I randomly get SEGV on pages like htt...
-
- 01:45 PM Revision c9af563e (git): * 2021-06-06 [ci skip]
-
-
09:56 AM Bug #17835: Net::HTTP should automatically add Accept-Encoding request header on HEAD request as well
- Opened: https://github.com/ruby/net-http/pull/26
-
- I'm negative about this, because of the overhead and complexity it would introduce while supporting both forms, and because it seems difficult (and annoying) to ever remove the current positional argument.
FWIW, some IDEs shows argume...
06/05/2021
-
- The test `TestEnv#test_huge_value` showed a difference between msvcrt and ucrt for x64-mingw32:
```
1) Failure:
TestEnv#test_huge_value [C:/CIFS/ruby-trunk/test/ruby/test_env.rb:472]:
Exception raised:
<#<Errno::EINVAL: Invalid ... -
11:33 AM Feature #17853: Add Thread#thread_id
- @nobu that looks great!
-
- ```diff
diff --git i/thread.c w/thread.c
index cbef44a9d4a..c627e5c8f2a 100644
--- i/thread.c
+++ w/thread.c
@@ -3456,6 +3456,14 @@ rb_thread_to_s(VALUE thread)
if (!NIL_P(target_th->name)) {
rb_str_catf(str, "@%"PRI... -
- @ko1 Could you reply to this?
I feel C code in the middle of Ruby code is not so great, and if moving C code to strings in Ruby is just to add `Primitive.attr! 'inline'` (like for `Float#zero?`),
it seems much simpler and nicer to ad... -
- When you want to make arguments self-explanatory, I wonder if it is not sufficient to write as follows:
```ruby
obj = Object.new
obj.respond_to?(:symbol, include_all = false)
# or
obj.methods(include_inherited = true)
```
-
- AFAIK, there was no such message as "Beginless range is experimental, and the behavior may change in future versions of Ruby!" in the first place. I am not responsible for this, but my feeling is that this feature is not quite unstable now.
-
06:48 AM Misc #17683: Current status of beginless range (experimental or not)
- Hello, is there anyone who is responsible for this? I need this information for the book I'm writing now.
-
04:57 AM Feature #17135 (Closed): Improve performance of Integer#size method
- Applied in changeset commit:git|3208a5df2dfb429752a130a36274464e9924cf44.
----------
Improve perfomance for Integer#size method [Feature #17135] (#3476)
* Improve perfomance for Integer#size method [Feature #17135]
* re-run ci
* Let ... -
- * Improve perfomance for Integer#size method [Feature #17135]
* re-run ci
* Let MJIT frame skip work for Integer#size
Co-authored-by: Takashi Kokubun <takashikkbn@gmail.com> - 03:28 AM Revision 033e76e7 (git): * 2021-06-05 [ci skip]
-
03:27 AM Revision 67ca2cce (git): Update bundled_gems
06/04/2021
-
- Some Ruby methods accept optional boolean arguments. This kind of parameter is known to be confusing since you cannot tell just looking at the method call what the parameter mean. For example:
```ruby
object.respond_to?(:symbol, fals... -
05:47 PM Feature #17763: Implement cache for cvars
- This is the script/app I used to measure https://github.com/k0kubun/railsbench
-
- I ran benchmarks using railsbench and the branch with the CVAR cache is a lot faster, 657 requests per second over the 615 requests per second on master.
Master:
```
$ RAILS_ENV=production INTERVAL=100 WARMUP=1 BENCHMARK=10000 rub... -
- If this is already fixed in 2.7.3, I think we can close this. We cannot change 2.7.1.
-
03:16 PM Bug #17937 (Closed): Segmentation fault in Enumerator#next on Apple M1, Mac OS Big Sur 11.2.2
- A simple code in the irb console fails:
```
irb(main):003:0> [1,2,3].to_enum.next
[BUG] Segmentation fault at 0x0000000004809020
ruby 2.7.1p83 (2020-03-31 revision a0c7c23c9c) [-darwin20]
...
-- Control frame information --------... -
- It is a known restriction that the code is parsed as `case (expression in 42)`.
Let me consider if it is possible to make that a priority. -
- I want to change the current condition to run Travis to improve developer's experience.
In my understanding, currently on ruby/ruby, Travis is executed at pull-request and push timing to only `master` or `ruby_N_N` branches by [the br... -
08:37 AM Feature #12075: some container#nonempty?
- I want to mention that some of the method names proposed so far would break symmetry.
Note that "empty" is a property of the container, not the elements. When you have `a = [1]`, what is questioned whether it is empty or not is `a`, but... -
- I like `Array#some?` and `Hash#some?`. It is so frequent to write `!ary.empty?`.
I think the main concern is only about `empty?`, not other predicate methods. So I'm skeptical about the need of the generalization like `ary.not(:empty?)`... -
- Also there is a problem with the proposed work around:
```
begin
rescue Exception => exception
raise
ensure
if exception
abort
else
commit
end
end
```
This only works provided you don't have any other `re... -
- Here are some examples of situations where `$!` is used inside `ensure` block. In every case, there is a potential buggy behaviour.
https://github.com/drbrain/net-http-persistent/blob/75574f2546a08aa2663b06a2e005bcf2ee304f13/lib/net/h... -
- Note that there are two related issues, for which the solutions conflict with each other:
* #9710: no __builtin_setjmp on x64-mingw
* #15348: use __builtin_setjmp on mingw64 -
- I have a similar problem with a recent 3.1.0dev version, in which the attached test case (named `x64-mingw32-callcc-test.rb`) dies silently.
I suppose that the code inserted by commit:a6ae274c3b06174401276fde2636f17720508532 (dated Ju... -
-
06/03/2021
- 08:57 PM Revision 9f3240d4 (git): * 2021-06-04 [ci skip]
-
- NUM_IN_PAGE(page->start) will sometimes return a 0 or a 1 depending on
how the alignment of the 40 byte slots work out. This commit uses the
NUM_IN_PAGE function to shift the bitmap down on the first bitmap plane.
Iterating on the first... -
12:18 PM Bug #17935: Backport 86c262541ad07528842d76dab4b9b34bd888d5f4 and 7e14762159643b4415e094f9d2a90afaf7994588 to Ruby 3.0
- ruby_3_0 2dd18df4a35a4b2dd0cf2dec7759898246fc6935 merged revision(s) 86c262541ad07528842d76dab4b9b34bd888d5f4,7e14762159643b4415e094f9d2a90afaf7994588.
-
-
- Sorry, I made a follow-up fix commit:7e14762159643b4415e094f9d2a90afaf7994588 as well.
-
- Please consider backporting commit:86c262541ad07528842d76dab4b9b34bd888d5f4 and commit:7e14762159643b4415e094f9d2a90afaf7994588 to ruby_3_0. This is a bug introduced by changes in Ruby 3.0 that causes SEGVs.
-
- ruby_3_0 9680ee97e0b3e87c0fc9a65c01de1ee50a1a178b merged revision(s) 1aac0e88193a82ed36b43e852c46414181b66455.
-
- Fix a race condition around mjit_recompile
This fixes SEGVs like https://github.com/ruby/ruby/runs/2715166621?check_suite_focus=true.
When mjit_recompile is called when mjit_compile is compiling the exact
same i... -
- Mark inlined ISeqs during MJIT compilation (#4539)
[Bug #17584]
---
common.mk | 6 ++++++
mjit.c | 18 +++++++++++-----
mjit_compile.c | 24 +++++++++++++++------
mjit_worke... -
-
10:15 AM Bug #17934: clipboard gem fail on Ruby 3.0.1
- In Windows x64, SIZE_T and HGLOBAL (and other handles) in Fiddle shoule be map to void*(64 bits), not uint(32 bits).
So gem clipboard V1.3.5 got a pointor failure.
Attachment fix this bug, just for Windows platform. - 08:47 AM Revision dc254120 (git): Expose `rb_fiber_transfer` and `rb_fiber_transfer_kw`.
-
- This is a follow-up of 86c262541ad07528842d76dab4b9b34bd888d5f4.
CRITICAL_SECTION_START/FINISH are not needed when it's called from an
MJIT worker.
Also, ZALLOC needs to be calloc because ZALLOC may trigger GC, which an
MJIT worker must... -
- Following non-special_const literals:
* T_REGEXP -
- Following non-special_const literals:
* T_BIGNUM
* T_FLOAT (non-flonum)
* T_RATIONAL
* T_COMPLEX -
-
-
- This fixes SEGVs like https://github.com/ruby/ruby/runs/2715166621?check_suite_focus=true.
When mjit_recompile is called when mjit_compile is compiling the exact
same iseq (and after it called mjit_capture_cc_entries), iseq->body->jit_u... -
- because a SEGV might happen on an MJIT worker. As you can clearly see
from `if (vm && ec) {`, ec is not guaranteed to exist here. -
- This changes "test/rubygems/test_case.rb" to "test/rubygems/helper.rb",
and "test/rubygems/test_utilities.rb" to "test/rubygems/utilities.rb".
The two files are a helper for tests, not test files. However, a file
starting with "test_" p... -
- Applied in changeset commit:git|28b481938b5c8211aad53ba82fe4ddd978ffc00f.
----------
Implemented some NilClass method in Ruby code is faster [Feature #17054] (#3366) -
06/02/2021
-
-
- Thanks for the discussion. I'll try to summarise my position briefly.
> Maybe we should open a new issue so that Timeout.timeout uses an Exception and not throw?
I agree Timeout should raise an exception internally and we should ch... -
- > A timeout error in my mind is more similar to a TERM signal, and Ruby handles that using an exception.
I think that's a good point.
I agree for similar things like Ctrl+C/SIGINT which is an `Interrupt` exception (< SignalException ... -
- Eregon (Benoit Daloze) wrote in #note-21:
> Thanks for the link.
> ...
Yes, it's used anytime you want to return a response immediately. I just used redirect as an example.
> If `redirect` would use a regular exception, and for ins... -
- Thanks for the link.
Is `throw` used for other things than `redirect`?
If `redirect` would use a regular exception, and for instance pass `[]` as the backtrace, it would probably be as efficient, but then gives the possibility to d... -
- Eregon (Benoit Daloze) wrote in #note-19:
> > Eregon (Benoit Daloze) throw should not abort the transaction, it should be committed.
> ...
Rails 7 is going to start treating non-local, non-exception exits as rolling back the transactio... -
- > I suspect that there is existing buggy code which has ensure blocks and checks $! which fails in the case I mentioned.
Typically that's an anti-pattern.
In 99% cases, the ensure code should do its logic regardless whether there is ... -
-
- We are only iterating over the eden heap so `heap_eden->total_pages`
contains the exact number of pages we need to allocate for.
`heap_allocated_pages` may contain pages in the tomb. -
- > Note that this failure was observed 4 days ago, at the build no #43628 - job number 509310540 first as far as I know. Since then, the failure rate is 4/35 = 11% so far.
Ah. No. The 11% is not correct. Skipping the 4 builds after my ... -
- Note that this failure was observed 4 days ago, at the build no `#43628` - job number `509310540` first as far as I know. Since then, the failure rate is 4/35 = 11% so far.
https://www.travis-ci.com/github/ruby/ruby/builds
-
- I am trying to skip the test on Travis ppc64le for now.
https://github.com/ruby/ruby/pull/4546
-
- The failure happened on Travis ppc64le again yesterday.
https://travis-ci.com/github/ruby/ruby/jobs/510153529#L2250
- 03:07 PM Revision 2a685da1 (git): * 2021-06-03 [ci skip]
-
03:07 PM Revision e451f0f6 (git): Fix `_MSC_VER` warnings
-
-
- Indeed.
And this seems to be another instance of LLP64 issues which wrongfully assume `sizeof(long) == sizeof(void *)`. Running on i386-mingw32 works as expected.
(Edited. LLP64, not ILP32.) -
- More likely to be an issue of https://github.com/janlelis/clipboard
The error means the clipboard gem tries to write to the NULL address, which is a bug. -
- Feel free to report it to https://github.com/ffi/ffi/issues. (BTW it is reproducible for me.)
-
- gem install clipboard (Which was used well in Ruby 2.XX):
Fetching clipboard-1.3.5.gem
Successfully installed clipboard-1.3.5
Parsing documentation for clipboard-1.3.5
Installing ri documentation for clipboard-1.3.5
Done installing ... -
-
-
- 10:38 AM Revision 2048dfc5 (git): Skip a `File.atime` test randomly failing on Travis ppc64le.
- See <https://bugs.ruby-lang.org/issues/17926>.
-
10:06 AM Misc #17932: 90s design (please lets move to 21st century)
- hsbt (Hiroshi SHIBATA) wrote in #note-4:
> I always welcome to your improvement at https://github.com/ruby/b.r-l.o .
Ok, I will. :) -
- xtkoba (Tee KOBAYASHI) wrote in #note-3:
> I do not understand what "search is broken" means. Could you explain what you tried to do? Searching in Ruby Issue Tracking System seems to just work for me.
You can see it in the screenshot... -
- > It seems you missed to write the result.
It's in the first comment: https://bugs.ruby-lang.org/issues/17763#note-1
Ruby master / Rails 6.1:
```
Warming up ———————————————————
logger 155.251k I/100ms
Calculatin... -
02:13 AM Feature #17763: Implement cache for cvars
- > We also benchmarked ActiveRecord::Base.logger since logger is a cvar and there are 63 modules in the inheritance chain. This is an example of a real-world improvement to Rails applications.
> ...
It seems you missed to write the resul... -
- Agreed, but I think @Dan0042's `Process._fork_` proposal is better.
-
- Since there seems to be no agreement on what a better and high-level API would look like, would it be reasonable to go back to the "Make Kernel.fork a delegator" proposal from <https://github.com/ruby/ruby/pull/4361>?
That patch seems... -
- It's been a way too much amount of ifdefs.
-
- and not for third-party libraries.
See: e6484a153038703447b50fcac26349249922ab28 -
- https://github.com/ruby/ruby/runs/2707566811#step:10:147
```
D:\a\ruby\ruby\src\mjit_worker.c(1212): warning C4090: 'function': different 'const' qualifiers
``` -
02:35 AM Revision 9137caaf (git): [ruby/rdoc] Add a dependency on psych gem 4.0.0 or newer
- https://github.com/ruby/rdoc/commit/ebe185c877
-
- https://github.com/ruby/rdoc/commit/b8d68fdd87