issue-ac7d14b8a47f54914bee2c26739ccda0b0ef32e5.yaml

--- !ditz.rubyforge.org,2008-03-06/issue 
title: check why duk.300 fails to complete a testcase run (two cases hang)
desc: ""
type: :task
component: duk
release: v0.9
reporter: sva <sami.vaarala@iki.fi>
status: :closed
disposition: :fixed
creation_time: 2013-10-24 00:28:37.142772 Z
references: []

id: ac7d14b8a47f54914bee2c26739ccda0b0ef32e5
log_events: 
- - 2013-10-24 00:28:37.286777 Z
  - sva <sami.vaarala@iki.fi>
  - created
  - ""
- - 2013-10-24 00:32:07.808836 Z
  - sva <sami.vaarala@iki.fi>
  - commented
  - |-
    One of the hanging cases is:
    
      test-builtin-global-uri-bruteforce.js
- - 2013-10-28 23:23:31.393441 Z
  - sva <sami.vaarala@iki.fi>
  - commented
  - |-
    The other hanging case is:
    
      test-numconv-tostring-fastpath.js
- - 2013-10-28 23:26:52.330534 Z
  - sva <sami.vaarala@iki.fi>
  - commented
  - |-
    When manually running profile 300 vs. profile 400, it's apparent that
    profile 300 gets slower and slower.  Since profile 300 has no refcounting,
    could this be a garbage collection issue?  Perhaps mark-and-sweeps are
    happening too often?
- - 2013-10-28 23:27:07.647729 Z
  - sva <sami.vaarala@iki.fi>
  - assigned to release v0.8 from v0.7
  - ""
- - 2013-11-26 22:06:08.256850 Z
  - sva <sami.vaarala@iki.fi>
  - assigned to release v0.9 from v0.8
  - ""
- - 2013-12-08 13:27:48.367589 Z
  - sva <sami.vaarala@iki.fi>
  - closed with disposition fixed
  - Fixed by mark-and-sweep stringtable leak fix.