| Issue | Fix Version | Assignee | Priority | Status | Resolution | Resolved In Build |
|---|---|---|---|---|---|---|
| JDK-2038425 | 1.4.0 | Robert Griesemer | P3 | Resolved | Fixed | beta |
While diagnosing 4377159, discovered a new crash (Windows NT 4.0, x86).
Verification code in InterpreterMacroAssembler::call_VM_leaf_base fails to
verify edi. It appears that the stack pointer may be off by one slot. Failure
is intermittent but always occurs in the same place.
To reproduce:
cd to %GammaBase%/Bugs/<BugID>
Have /usr/local/java mounted on %LOCAL_JAVA%.
Have a java_g with the debug client compiler build in JDK pointed to by
%JAVA_HOME%.
doit %JAVA_HOME% %LOCAL_JAVA%\jck1.2
Note that the test case (jck12a012,java) has been cut down somewhat. Cutting it
down further seems to reduce the probability that the failure will occur. Have
tried running this with -XX:+ShowSafepointMsgs -XX:+Verbose and failure occurs
without any output from the safepoint code, so believe this is unrelated to
recent changes there.
Verification code in InterpreterMacroAssembler::call_VM_leaf_base fails to
verify edi. It appears that the stack pointer may be off by one slot. Failure
is intermittent but always occurs in the same place.
To reproduce:
cd to %GammaBase%/Bugs/<BugID>
Have /usr/local/java mounted on %LOCAL_JAVA%.
Have a java_g with the debug client compiler build in JDK pointed to by
%JAVA_HOME%.
doit %JAVA_HOME% %LOCAL_JAVA%\jck1.2
Note that the test case (jck12a012,java) has been cut down somewhat. Cutting it
down further seems to reduce the probability that the failure will occur. Have
tried running this with -XX:+ShowSafepointMsgs -XX:+Verbose and failure occurs
without any output from the safepoint code, so believe this is unrelated to
recent changes there.
- backported by
-
JDK-2038425 sin/cos related test crashes in call_VM_leaf_base
-
- Resolved
-
- relates to
-
JDK-4377159 HS1.3.1(win) intermittently crashes at os_win32.cpp, 837
-
- Closed
-