diff options
| author | NARUSE, Yui <nurse@users.noreply.github.com> | 2024-03-21 10:45:01 +0900 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-03-21 01:45:01 +0000 |
| commit | a63e979853783601a60228b45741f8b3776e5507 (patch) | |
| tree | 8396278af85906d600486360bfb7e466aaef6161 /test/ruby | |
| parent | a406c516685f1950269c4e43d13cc748f0bfbc06 (diff) | |
merge revision(s) d19d683a354530a27b4cbb049223f8dc70c75849,de1a586ecc2ee7f465f0c0a69291054136a3a819: [Backport #20250] (#10308)
rb_obj_setup: do not copy RUBY_FL_SEEN_OBJ_ID
[Bug #20250]
We're seting up a new instance, so it never had an associated
object_id.
proc.c: get rid of `CLONESETUP`
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
[Bug #20253]
All the way down to Ruby 1.9, `Proc`, `Method`, `UnboundMethod`
and `Binding` always had their own specific clone and dup routine.
This caused various discrepancies with how other objects behave
on `dup` and `clone. [Bug #20250], [Bug #20253].
This commit get rid of `CLONESETUP` and use the the same codepath
as all other types, so ensure consistency.
NB: It's still not accepting the `freeze` keyword argument on `clone`.
Co-Authored-By: Étienne Barrié <etienne.barrie@gmail.com>
Diffstat (limited to 'test/ruby')
| -rw-r--r-- | test/ruby/test_clone.rb | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/test/ruby/test_clone.rb b/test/ruby/test_clone.rb index 216eaa39d2..775c9ed848 100644 --- a/test/ruby/test_clone.rb +++ b/test/ruby/test_clone.rb @@ -73,6 +73,13 @@ class TestClone < Test::Unit::TestCase assert_equal(cloned_obj.instance_variable_get(:@a), 1) end + def test_proc_obj_id_flag_reset + # [Bug #20250] + proc = Proc.new { } + proc.object_id + proc.clone.object_id # Would crash with RUBY_DEBUG=1 + end + def test_user_flags assert_separately([], <<-EOS) # |