Project

General

Profile

Actions

Bug #10618

closed

TestWeakRef#test_repeated_object_leak fails on ARM

Added by vo.x (Vit Ondruch) about 10 years ago. Updated almost 10 years ago.

Status:
Closed
Target version:
-
ruby -v:
ruby -v: ruby 2.2.0dev (2014-12-17 trunk 48879) [armv7hl-linux]
[ruby-core:66949]

Description

I observer following error on Fedora's ARM builders:

  1) Error:
TestWeakRef#test_repeated_object_leak:
Timeout::Error: execution of assert_no_memory_leak expired
    /builddir/build/BUILD/ruby-2.2.0-r48879/test/test_weakref.rb:66:in `test_repeated_object_leak'

This issue was very likely introduced by r48820. Less iterations or longer timeout might help probably.


Related issues 1 (0 open1 closed)

Related to Ruby master - Bug #10537: Repeated creation and garbage collection of WeakRef instances against a single object leaks memoryClosed11/24/2014Actions

Updated by normalperson (Eric Wong) about 10 years ago

Is 60 seconds enough? Can 30 work? Default is only 10 seconds.

diff --git a/test/test_weakref.rb b/test/test_weakref.rb
index 36b5d5b..d21b431 100644
--- a/test/test_weakref.rb
+++ b/test/test_weakref.rb
@@ -63,7 +63,7 @@ class TestWeakRef < Test::Unit::TestCase

def test_repeated_object_leak
  bug10537 = '[ruby-core:66428]'
  • assert_no_memory_leak(%w(-rweakref), '', <<-'end;', bug10537)
  • assert_no_memory_leak(%w(-rweakref), '', <<-'end;', bug10537, timeout: 60)
    a = Object.new
    150_000.times { WeakRef.new(a) }
    end;

Updated by vo.x (Vit Ondruch) about 10 years ago

I ran 3 builds with following results:

TestWeakRef#test_repeated_object_leak = 19.71 s = .
TestWeakRef#test_repeated_object_leak = 20.04 s = .
TestWeakRef#test_repeated_object_leak = 19.88 s = .

Actions #3

Updated by Anonymous almost 10 years ago

  • Status changed from Open to Closed
  • % Done changed from 0 to 100

Applied in changeset r48900.


test/test_weakref.rb: increase timeout [Bug #10618]

  • test/test_weakref.rb (test_repeated_object_leak): increase timeout
    [Bug #10618]

Updated by normalperson (Eric Wong) almost 10 years ago

Thanks, I just bumped to 60s in r48900 (and r48902 for ChangeLog)
in case there are slower machines.

Btw, I'm curious: how long does "test-all" take for your ARM machines?

Updated by vo.x (Vit Ondruch) almost 10 years ago

You can peek yourself into Fedora's builder [1]. The build.log for this particular build [2] says:

Finished tests in 1428.934639s, 11.1552 tests/s, 1757.1049 assertions/s.
15940 tests, 2510788 assertions, 0 failures, 0 errors, 38 skips
ruby -v: ruby 2.2.0rc1 (2014-12-18 trunk 48887) [armv7hl-linux]
check succeeded

[1] http://koji.fedoraproject.org/koji/taskinfo?taskID=8437177
[2] https://kojipkgs.fedoraproject.org//work/tasks/7180/8437180/build.log

Updated by vo.x (Vit Ondruch) almost 10 years ago

  • Backport changed from 2.0.0: UNKNOWN, 2.1: UNKNOWN to 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: REQUIRED

Updated by naruse (Yui NARUSE) almost 10 years ago

  • Backport changed from 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: REQUIRED to 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: DONE

ruby_2_2 r48949 merged revision(s) 48900,48902.

Actions #8

Updated by usa (Usaku NAKAMURA) almost 10 years ago

  • Related to Bug #10537: Repeated creation and garbage collection of WeakRef instances against a single object leaks memory added
Actions #9

Updated by usa (Usaku NAKAMURA) almost 10 years ago

  • Backport changed from 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: DONE to 2.0.0: REQUIRED, 2.1: REQUIRED, 2.2: DONE

Updated by usa (Usaku NAKAMURA) almost 10 years ago

  • Backport changed from 2.0.0: REQUIRED, 2.1: REQUIRED, 2.2: DONE to 2.0.0: DONTNEED, 2.1: DONTNEED, 2.2: DONE
Actions

Also available in: Atom PDF

Like0
Like0Like0Like0Like0Like0Like0Like0Like0Like0Like0