JDK-4614097 : KLA-Tencor Cougar Application crash VM with -hotspot
  • Type: Bug
  • Component: hotspot
  • Sub-Component: runtime
  • Affected Version: 1.3.1,1.3.1_02
  • Priority: P4
  • Status: Closed
  • Resolution: Fixed
  • OS: other,windows_nt
  • CPU: x86
  • Submitted: 2001-12-14
  • Updated: 2012-10-08
  • Resolved: 2002-04-09
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
Other
1.3.1_04 04Fixed
Related Reports
Duplicate :  
JDK-4517252 - Hotspot crashes
Relates :  
JDK-4631621 - KLA-Tencor Cougar application causes VM crash with -server
Description
JVM crashes while running "Cougar" atop JRE 1.3.1_02 on Windows 2000.
"Cougar" is an application developed by KLA-Tencor and being evaluated 
for inclusion in the Java Software BigApps testing program.

Yumin Qi of CTE has stack trace and other postmortem details.  He 
reports that preliminary analysis suggests a GC problem.
Comments
CONVERTED DATA BugTraq+ Release Management Values COMMIT TO FIX: 1.3.1_04 FIXED IN: 1.3.1_04 INTEGRATED IN: 1.3.1_04
14-06-2004
EVALUATION The bug for the -hotspot is a bad oopmap at a backward branch safepoint. What's happening is that one of the locals inside the loop is being assigned to a register for the lifetime of the loop but it's not alive at the backward branch. The code which handles locals that are assigned to registers was ignoring the liveness analysis being done by our oopmap generator and so it was putting the register in the oop map for the backward branch. The fix to modify that code to take into account the results of the ciLocalMap in determining whether a loop local in a register is still alive. This bug has been in c1 since the beginning but it only affects the intel side since the sparc side uses a slightly different mechanism which correctly consults the ciLocalMap. 1.4 is unaffected by this bug since the code has been completely rewritten and is modeled after the sparc implementation. This bug is completely differnt than whatever problem was affecting -server. ###@###.### 2002-01-31
31-01-2002