| Age | Commit message (Collapse) | Author |
|---|
|
* Make sure to always use the right `warn`
* lib/bundled_gems.rb: more reliable caller detection
The `2` skipped frames went out of sync and now it should be `3`.
Rather than just update the offset, we can implement a way that
is adaptative as long as all require decorators are also called require.
Also we should compute the corresponding `uplevel` otherwise the
warning will still point decorators.
Co-authored-by: "Hiroshi SHIBATA" <[email protected]>
* Warn ostruct for Ruby 3.5
* Warn pstore for Ruby 3.5
* Mark rdoc as bundled gems at Ruby 3.5
* Warn to use win32ole without Gemfile for Ruby 3.5
* EXACT list is mostly same as SINCE list on bundled gems.
* Mark to warn fiddle as bundled gems for Ruby 3.5
* Mark to warn logger as bundled gems for Ruby 3.5
* We should use uplevel:2 in another case.
Like the following scenario with bootsnap, that frames are same or smaller than frame_to_skip(=3).
---
"/Users/hsbt/.local/share/rbenv/versions/3.3-dev/lib/ruby/3.3.0/bundled_gems.rb:69:in `block (2 levels) in replace_require'"
"/Users/hsbt/.local/share/gem/gems/bootsnap-1.18.4/lib/bootsnap/load_path_cache/core_ext/kernel_require.rb:30:in `require'"
"test_warn_bootsnap.rb:11:in `<main>'"
---
* Delete unnecessary rubocop disable comment
* Show correct script name with sub-feature case
* Skip to show script name with using ruby -r option
* Don't show script name when bundle exec and call ruby script directly.
* Pick word fix from 34adc07372c10170b8ca36111d216cbd8e4699be
---------
Co-authored-by: David Rodríguez <[email protected]>
Co-authored-by: Jean Boussier <[email protected]>
Co-authored-by: Kentaro Takeyama <[email protected]>
|
|
[Bug #20633] Re-initialize vm->ractor.sched.lock after fork
Previously under certain conditions it was possible to encounter a
deadlock in the forked child process if ractor.sched.lock was held.
Co-authored-by: Nathan Froyd <[email protected]>
|
|
|
|
[Bug #20668]
The `dest` node is assumed to be a `CDECL`, so overwriting it with
a `LIT` cause a crash on the next iteration.
Co-authored-by: Jean Boussier <[email protected]>
|
|
* Merge RubyGems-3.5.12 and Bundler-2.5.12
* Merge RubyGems-3.5.13 and Bundler-2.5.13
* Merge RubyGems-3.5.14 and Bundler-2.5.14
* Merge RubyGems-3.5.15 and Bundler-2.5.15
* Merge RubyGems-3.5.16 and Bundler-2.5.16
|
|
* Fix floor when ndigits is large
[Bug #20654]
This commit fixes Integer#floor and Float#floor when the number is
negative and ndigits is large such that 10**ndigits is a bignum.
Previously, it would return 0 in such cases. However, this would cause
unexpected behaviour such as:
puts -1.floor(-5) # => -100000
puts -1.floor(-10) # => -10000000000
puts -1.floor(-20) # => 0
This commit changes the last result so that it will return
-100000000000000000000.
* Fix ceil when ndigits is large
[Bug #20654]
This commit fixes Integer#ceil and Float#ceil when the number is
negative and ndigits is large such that 10**ndigits is a bignum.
Previously, it would return 0 in such cases. However, this would cause
unexpected behaviour such as:
puts 1.ceil(-5) # => 100000
puts 1.ceil(-10) # => 10000000000
puts 1.ceil(-20) # => 0
This commit changes the last result so that it will return
100000000000000000000.
|
|
(#11255)
Fix memory leak in String#start_with? when regexp times out
[Bug #20653]
This commit refactors how Onigmo handles timeout. Instead of raising a
timeout error, onig_search will return a ONIGERR_TIMEOUT which the
caller can free memory, and then raise a timeout error.
This fixes a memory leak in String#start_with when the regexp times out.
For example:
regex = Regexp.new("^#{"(a*)" * 10_000}x$", timeout: 0.000001)
str = "a" * 1000000 + "x"
10.times do
100.times do
str.start_with?(regex)
rescue
end
puts `ps -o rss= -p #{$$}`
end
Before:
33216
51936
71152
81728
97152
103248
120384
133392
133520
133616
After:
14912
15376
15824
15824
16128
16128
16144
16144
16160
16160
|
|
|
|
Fix memory leak in Regexp capture group when timeout
[Bug #20650]
The capture group allocates memory that is leaked when it times out.
For example:
re = Regexp.new("^#{"(a*)" * 10_000}x$", timeout: 0.000001)
str = "a" * 1000000 + "x"
10.times do
100.times do
re =~ str
rescue Regexp::TimeoutError
end
puts `ps -o rss= -p #{$$}`
end
Before:
34688
56416
78288
100368
120784
140704
161904
183568
204320
224800
After:
16288
16288
16880
16896
16912
16928
16944
17184
17184
17200
|
|
bundled_gems.rb: Add a fast path
[Bug #20641] `Gem::BUNDLED_GEMS.warning?` adds a lot of extra
work on top of `require`. When the call end up atually loading code
the overhead is somewhat marginal.
However it's not uncommon for code to go some late `require` in some
paths, so it's expected that calling `require` with something already
required is somewhat fast, and `bundled_gems.rb` breaks this assumption.
To avoid this, we can have a fast path that in most case allow to
short-circuit all the heavy computations. If we extract the feature
basename and it doesn't match any of the bundled gems we care about
we can return very early.
With this change `require 'date'` is now only 1.33x slower on Ruby
3.3.3, than it was on Ruby 3.2.2, whereas before this change it
was at least 100x slower.
Co-authored-by: Jean Boussier <[email protected]>
|
|
[Bug #20633] Fix the condition for `atomic_signal_fence`
`AC_CHECK_DECLS` defines `HAVE_DECL_SYMBOL` to 1 if declared, 0
otherwise, not undefined.
Co-authored-by: kimuraw (Wataru Kimura) <[email protected]>
|
|
When the registerred unblock function is called, it should retry
the cancelled blocking function if possible after checkints.
For example, `SIGCHLD` can cancel this method, but it should not
raise any exception if there is no trap handlers.
The following is repro-code:
```ruby
require 'socket'
PN = 10_000
1000000.times{
p _1
PN.times{
fork{
sleep rand(0.3)
}
}
i = 0
while i<PN
cpid = Process.wait -1, Process::WNOHANG
if cpid
# p [i, cpid]
i += 1
end
begin
TCPServer.new(nil, 0).close
rescue
p $!
exit!
end
end
}
```
|
|
|
|
clear `kw_flag` if given hash is nil
https://bugs.ruby-lang.org/issues/20570 is caused I missed to
clear the `kw_flag` even if `keyword_hash` is nil.
|
|
[Bug #20239] Fix overflow at down-casting
|
|
[Backport #20592]
[Bug #20592] Fix segfault when sending NULL to freeaddrinfo
On alpine freeaddrinfo does not accept NULL pointer
Fix dangling `else`
|
|
[Bug #20468] Fix safe navigation in `for` variable
|
|
[Bug #20562] Categorize `RUBY_FREE_AT_EXIT` warning as experimental
|
|
|
|
Update macos runners with latest environments.
* Use macos-14 instead of macos-arm-oss
* Removed macos-11 and added macos-13
|
|
safe profiling (#11090)
**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)
can safely sample from the signal handler.
This is a backport of #11036 to Ruby 3.3 .
**Motivation:**
The `vm_push_frame` was specifically tweaked in
https://github.com/ruby/ruby/pull/3296 to initialize the a frame
before updating the `cfp` pointer.
But since there's nothing stopping the compiler from reordering
the initialization of a frame (`*cfp =`) with the update of the cfp
pointer (`ec->cfp = cfp`) we've been hesitant to rely on this on
the Datadog profiler.
In practice, after some experimentation + talking to folks, this
reordering does not seem to happen.
But since modern compilers have a way for us to exactly tell them
not to do the reordering (`atomic_signal_fence`), this seems even
better.
I've actually extracted `vm_push_frame` into the "Compiler Explorer"
website, which you can use to see the assembly output of this function
across many compilers and architectures: https://godbolt.org/z/3oxd1446K
On that link you can observe two things across many compilers:
1. The compilers are not reordering the writes
2. The barrier does not change the generated assembly output
(== has no cost in practice)
**Additional Notes:**
The checks added in `configure.ac` define two new macros:
* `HAVE_STDATOMIC_H`
* `HAVE_DECL_ATOMIC_SIGNAL_FENCE`
Since Ruby generates an arch-specific `config.h` header with
these macros upon installation, this can be used by profilers
and other libraries to test if Ruby was compiled with the fence enabled.
**How to test the change?**
As I mentioned above, you can check https://godbolt.org/z/3oxd1446K
to confirm the compiled output of `vm_push_frame` does not change
in most compilers (at least all that I've checked on that site).
|
|
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-House <[email protected]>
|
|
Fix corruption of encoding name string
[Bug #20595]
enc_set_default_encoding will free the C string if the encoding is nil,
but the C string can be used by the encoding name string. This will cause
the encoding name string to be corrupted.
Consider the following code:
Encoding.default_internal = Encoding::ASCII_8BIT
names = Encoding.default_internal.names
p names
Encoding.default_internal = nil
p names
It outputs:
["ASCII-8BIT", "BINARY", "internal"]
["ASCII-8BIT", "BINARY", "\x00\x00\x00\x00\x00\x00\x00\x00"]
Co-authored-by: Matthew Valentine-House <[email protected]>
|
|
gems (#11006)
* Try to load original gemspec from `.bundle/gems/foo-x.y.z/foo.gemspec`.
`.bundle/specification/foo-x.y.z.gemspec` may be changed our toolchain
* Try to find gemspec from `.bundle/specifications
* Adjust indent
|
|
|
|
[Bug #20585]
This was changed in 36a06efdd9f0604093dccbaf96d4e2cb17874dc8 because
`String.new(1024)` would end up allocating `1025` bytes, but the problem
with this change is that the caller may be trying to right size a String.
So instead, we should just better document the behavior of `capacity:`.
Co-authored-by: Jean Boussier <[email protected]>
|
|
Don't call `Warning.warn` unless the category is enabled
The warning category should be enabled if we want to call
`Warning.warn`.
This commit speeds up the following benchmark:
```ruby
eval "def test; " +
1000.times.map { "' '.chomp!" }.join(";") + "; end"
def run_benchmark count
i = 0
while i < count
start = Process.clock_gettime(Process::CLOCK_MONOTONIC)
yield
ms = Process.clock_gettime(Process::CLOCK_MONOTONIC) - start
puts "itr ##{i}: #{(ms * 1000).to_i}ms"
i += 1
end
end
run_benchmark(25) do
250.times do
test
end
end
```
On `master` this runs at about 92ms per iteration. With this patch, it
is 7ms per iteration.
[Bug #20573]
|
|
`yjit-bindgen` isn't run to build Ruby releases at all, but people might
be running security scanners on the source tarball. Upgrade this
dependency to calm the scanners.
Backport of <https://github.com/ruby/ruby/pull/9652>.
See: <https://github.com/ruby/ruby/pull/10980>
|
|
|
|
Co-authored-by: Takashi Kokubun <[email protected]>
|
|
Some of the places in Redmine (e.g. Associated revisions) print
revisions using only 8 characters. Even when I copied a revision from
there, I want to prepend commit: in the message.
|
|
[Bug #20521] ripper: Clean up strterm
|
|
[Bug #20499] Use Xcode owned tools for Xcode clang
Xcode has its own version tools that may be incompatible with genuine
LLVM tools, use the tools in the same directory.
|
|
compile.c: use putspecialobject for RubyVM::FrozenCore
[Bug #20569]
`putobject RubyVM::FrozenCore`, is not serializable, we
have to use `putspecialobject VM_SPECIAL_OBJECT_VMCORE`.
Co-authored-by: Jean Boussier <[email protected]>
|
|
[Bug #20280]
Backport of #10014.
|
|
[Bug #20180]
Backports #9624.
|
|
If you start Ruby with `--yjit-disable`, the `+YJIT` shouldn't be
added until `RubyVM::YJIT.enable` is actually called. Otherwise
it's confusing in crash reports etc.
Co-authored-by: Jean Boussier <[email protected]>
|
|
Improve YJIT performance warning regression test
[Bug #20522]
|
|
Co-authored-by: Nobuyoshi Nakada <[email protected]>
|
|
[Bug #20517] Make a multibyte character one token at meta escape
|
|
(#10911)
Do not emit shape transition warnings when YJIT is compiling
[Bug #20522]
If `Warning.warn` is redefined in Ruby, emitting a warning would invoke
Ruby code, which can't safely be done when YJIT is compiling.
Co-authored-by: Jean Boussier <[email protected]>
Co-authored-by: Takashi Kokubun <[email protected]>
|
|
This is a backport of 6c8ae44a388e5c03b7db90376af3652007b574e8 with a
test tailored to crash the 3.3.x branch (from GH-10904).
Previously, we read the last element array even when the array was
empty, doing an out-of-bounds access. This sometimes caused a SEGV.
[Bug #20496]
|
|
338eb0065bd81ba8ae8b9402abc94804a24594cc, ac636f5709feb1d9d7a0c46a86be153be765cf21: [Backport #20516]
Revert "Rollback to released version numbers of stringio and strscan"
This reverts commit 6a79e53823e328281b9e9eee53cd141af28f8548.
[ruby/strscan] StringScanner#captures: Return nil not "" for unmached capture (https://github.com/ruby/strscan/pull/72)
fix https://github.com/ruby/strscan/issues/70
If there is no substring matching the group (s[3]), the behavior is
different.
If there is no substring matching the group, the corresponding element
(s[3]) should be nil.
```
s = StringScanner.new('foobarbaz') #=> #<StringScanner 0/9 @ "fooba...">
s.scan /(foo)(bar)(BAZ)?/ #=> "foobar"
s[0] #=> "foobar"
s[1] #=> "foo"
s[2] #=> "bar"
s[3] #=> nil
s.captures #=> ["foo", "bar", ""]
s.captures.compact #=> ["foo", "bar", ""]
```
```
s = StringScanner.new('foobarbaz') #=> #<StringScanner 0/9 @ "fooba...">
s.scan /(foo)(bar)(BAZ)?/ #=> "foobar"
s[0] #=> "foobar"
s[1] #=> "foo"
s[2] #=> "bar"
s[3] #=> nil
s.captures #=> ["foo", "bar", nil]
s.captures.compact #=> ["foo", "bar"]
```
https://docs.ruby-lang.org/ja/latest/method/MatchData/i/captures.html
```
/(foo)(bar)(BAZ)?/ =~ "foobarbaz" #=> 0
$~.to_a #=> ["foobar", "foo", "bar", nil]
$~.captures #=> ["foo", "bar", nil]
$~.captures.compact #=> ["foo", "bar"]
```
* StringScanner#captures is not yet documented.
https://docs.ruby-lang.org/ja/latest/class/StringScanner.html
https://github.com/ruby/strscan/commit/1fbfdd3c6f
[ruby/strscan] Bump version
https://github.com/ruby/strscan/commit/d6f97ec102
|
|
0301473fb523c71d8cdc4966971f31f502001185, 874e9fc34d728f8e2444d15aa6759befd217c464, 7f0e26b7f99bf76408569892ce20318501f74729: [Backport #20516]
Clear runtime dependencies if default gems is specified.
The current build system uses runtime dependencies from only
`.bundle` directory. We shouldn't install runtime dependencies
from rubygems.org when `make test-bundled-gems` is invoked.
Fixed dependencies list format
Don't need to remove ruby2_keywords dependency from drb
Re-use strscan with ruby repo
|
|
so that they are linked correctly on GitHub
|
|
Update bundled_gems
|
|
Do not pollute toplevel namespace
|
|
test_bignum: defined? returns String (#10880)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
didn't verify the test is working properly due to mistaken auto-merge… [Bug #20515]
bug: https://bugs.ruby-lang.org/issues/20515
follow-up: 22e4eeda6561693367fc7a00b92b90f46b09cabd
follow-up: https://github.com/ruby/ruby/pull/10875
|
|
22e4eeda6561693367fc7a00b92b90f46b09cabd,1ab7c412d2e3880a7ad233c32e93961888f8145c: [Backport #20515]
ci: Test whether GMP is working in compilers.yml (#10875)
Avoid reoccurence of [Bug #20515]
Requires https://github.com/ruby/ruby/pull/10876 since 18eaf0be905e3e251423b42d6f4e56b7cae1bc3b
bug: https://bugs.ruby-lang.org/issues/20515
RUBY_CHECK_HEADER didn't define HAVE_{header-file} (#10876)
--with-gmp is not working at all because HAVE_GMP_H
was missing since 18eaf0be90. [Bug #20515]
bug: https://bugs.ruby-lang.org/issues/20515
follow-up: https://bugs.ruby-lang.org/issues/20494
follow-up: 18eaf0be905e3e251423b42d6f4e56b7cae1bc3b
follow-up: https://github.com/ruby/ruby/pull/10805
|
|
055613fd868a8c94e43893f8c58a00cdd2a81f6d,127d7a35df10ee2bc99f44b888972b2c5361d84f,e2a9b87126d59e4766479a7aa12cf7a648f46506: [Backport #20447]
Fix pointer incompatiblity
Since the subsecond part is discarded, WIDEVAL to VALUE conversion is
needed.
Some functions are not used when `THREAD_MODEL=none`
`rb_thread_sched_destroy` is not used now at all
|
