JDK-4295347 : jdk1.3.0P HotSpot fails to pass api/java_lang/Thread tests,
  • Type: Bug
  • Component: hotspot
  • Sub-Component: compiler
  • Affected Version: 1.3.0
  • Priority: P1
  • Status: Closed
  • Resolution: Duplicate
  • OS: solaris_7
  • CPU: sparc
  • Submitted: 1999-11-30
  • Updated: 2021-11-09
  • Resolved: 1999-12-02
Related Reports
Duplicate :  
JDK-4295241 - 1.3FCS-P,HotSpot,Solaris: Object.wait(timeout) waits for incorrect elapsed time
Description

Name: akC45999			Date: 11/30/99



 The following JCK api tests fails (the tests hang up and 
 are interrupted) running with HotSpot, jdk1.3.0P on Solaris:
 api/java_lang/Thread/join01/join0101/join0101.html
 api/java_lang/Thread/join02/join0201/join0201.html

 However, the tests pass with jdk1.3.0O HotSpot on Solaris and WinNT.

 The tests also pass with HotSpot on WinNT, jdk1.3.0P.
 The tests pass with classic VM of jdk1.3.0O, jdk1.3.0P, 
 Solaris, WinNT.
 
 The tests exist in JCK116, JCK122, they were never excluded.

 ======================================================================
======================================================================
Comments
EVALUATION y.s.ramakrishna@eng 1999-12-01: Both tests consistently pass for me, using build P of the JDK, and the latest 1.3-ref and hotspot_baseline builds. With the VM that was packaged with 1.3P JDK, both tests consistently fail.] I am trying to determine (at a low priority) the reason for the failure in the old VM, and the change that resulted in the bug getting fixed in the more recent VM. y.s.ramakrishna@eng 1999-12-01: The cause is the same as that for bug 4295241: a problem with Object.wait(), which is waiting always for 8 minutes. So the "hang" in the description is really an 8 minute wait (times 4). So your test will pass after 24+ minutes... Oh well. Will dup this bug to 4295421 and update that bug (also no longer present in the newest VM) with what i learn as being the cause of the bug and its mysterious disappearance.
11-06-2004