-
Bug
-
Resolution: Fixed
-
P2
-
6
This test started failing in the 04-07-05 nightly, on various x86
platforms when run with -Xcomp (both client and server).
service_hs_baseline was resynced with main/baseline on 2005.04.07.
c2_baseline was pushed to main/baseline on 2005.04.07 and no
other changes to main/baseline were made. That is why this is being
reported against c2.
The failure mode is this on unix:
This test failed the following assertion on Linux IA32 Client VM
(machine chandrila), Solaris X86 Client VM (machine tictoc),
and Solaris X86 Server VM (machine kashyyyk):
Internal Error (src/share/vm/runtime/handles.cpp, 16)
assert(SharedSkipVerify || obj->is_oop(),"sanity check")
and on Win32, we get an EXCEPTION_ACCESS_VIOLATION. hs_err says:
Instructions: (pc=0x081e1cbb)
0x081e1cab: 75 04 8b cb eb 20 8b 44 24 34 85 c0 74 02 8b 00
0x081e1cbb: 8b 94 b8 08 01 00 00 8d 4c 24 3c 52 56 e8 c3 4f
;; 081e1cab 75 04 jne 0x081e1cb1
;; 081e1cad 8b cb mov %ebx,%ecx
;; 081e1caf eb 20 jmp 0x081e1cd1
;; 081e1cb1 8b 44 24 34 mov 0x34(%esp,1),%eax
;; 081e1cb5 85 c0 test %eax,%eax
;; 081e1cb7 74 02 je 0x081e1cbb
;; 081e1cb9 8b 00 mov (%eax),%eax
;; ---------------
;; 081e1cbb 8b 94 b8 08 01 00 00 mov 0x108(%eax,%edi,4),%edx
;; 081e1cc2 8d 4c 24 3c lea 0x3c(%esp,1),%ecx
;; 081e1cc6 52 push %edx
;; 081e1cc7 56 push %esi
;; 081e1cc8 e8 c3 4f ff ff call 0x081d6c90
;;
Stack: [0x00030000,0x00070000), sp=0x0006f978, free space=254k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
V [jvm.dll+0x1e1cbb];; ?runtime_resolve_virtual_method@LinkResolver@@CAXAAVCallInfo@@VmethodHandle@@VKlassHandle@@VHandle@@2HPAVThread@@@Z+0xeb
Java frames: (J=compiled Java code, j=interpreted, Vv=VM code)
j PopAsynchronousTarg.main([Ljava/lang/String;)V+26
v ~StubRoutines::call_stub
See attached hs_err file.
I seem to be able to reproduce this all the time on Win32, and only
rarely on solx86/linx86.
To reproduce, you can do this to create a partial SDK workspace that
contains the jtreg tests:
/net/jano.sfbay/export/disk20/serviceability/tools/bin/getjtregtests -jdi
You can then do
cd svcTests/test/com/sun/jdi
runregress -no PopAsynchronousTest.java -jdk <pathname to your jdk>
runregress is in the same dir as getjtregtests.
###@###.### 2005-04-18 19:17:46 GMT
This bug affects the following test:
com/sun/jdi/PopAsynchronousTest.java fails on x86
###@###.### 2005-04-19 00:04:37 GMT
platforms when run with -Xcomp (both client and server).
service_hs_baseline was resynced with main/baseline on 2005.04.07.
c2_baseline was pushed to main/baseline on 2005.04.07 and no
other changes to main/baseline were made. That is why this is being
reported against c2.
The failure mode is this on unix:
This test failed the following assertion on Linux IA32 Client VM
(machine chandrila), Solaris X86 Client VM (machine tictoc),
and Solaris X86 Server VM (machine kashyyyk):
Internal Error (src/share/vm/runtime/handles.cpp, 16)
assert(SharedSkipVerify || obj->is_oop(),"sanity check")
and on Win32, we get an EXCEPTION_ACCESS_VIOLATION. hs_err says:
Instructions: (pc=0x081e1cbb)
0x081e1cab: 75 04 8b cb eb 20 8b 44 24 34 85 c0 74 02 8b 00
0x081e1cbb: 8b 94 b8 08 01 00 00 8d 4c 24 3c 52 56 e8 c3 4f
;; 081e1cab 75 04 jne 0x081e1cb1
;; 081e1cad 8b cb mov %ebx,%ecx
;; 081e1caf eb 20 jmp 0x081e1cd1
;; 081e1cb1 8b 44 24 34 mov 0x34(%esp,1),%eax
;; 081e1cb5 85 c0 test %eax,%eax
;; 081e1cb7 74 02 je 0x081e1cbb
;; 081e1cb9 8b 00 mov (%eax),%eax
;; ---------------
;; 081e1cbb 8b 94 b8 08 01 00 00 mov 0x108(%eax,%edi,4),%edx
;; 081e1cc2 8d 4c 24 3c lea 0x3c(%esp,1),%ecx
;; 081e1cc6 52 push %edx
;; 081e1cc7 56 push %esi
;; 081e1cc8 e8 c3 4f ff ff call 0x081d6c90
;;
Stack: [0x00030000,0x00070000), sp=0x0006f978, free space=254k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
V [jvm.dll+0x1e1cbb];; ?runtime_resolve_virtual_method@LinkResolver@@CAXAAVCallInfo@@VmethodHandle@@VKlassHandle@@VHandle@@2HPAVThread@@@Z+0xeb
Java frames: (J=compiled Java code, j=interpreted, Vv=VM code)
j PopAsynchronousTarg.main([Ljava/lang/String;)V+26
v ~StubRoutines::call_stub
See attached hs_err file.
I seem to be able to reproduce this all the time on Win32, and only
rarely on solx86/linx86.
To reproduce, you can do this to create a partial SDK workspace that
contains the jtreg tests:
/net/jano.sfbay/export/disk20/serviceability/tools/bin/getjtregtests -jdi
You can then do
cd svcTests/test/com/sun/jdi
runregress -no PopAsynchronousTest.java -jdk <pathname to your jdk>
runregress is in the same dir as getjtregtests.
###@###.### 2005-04-18 19:17:46 GMT
This bug affects the following test:
com/sun/jdi/PopAsynchronousTest.java fails on x86
###@###.### 2005-04-19 00:04:37 GMT
- relates to
-
JDK-6260309 Regression: nsk/jvmti/PopFrame/popframe003 fails on x86 with -Xcomp
- Resolved
-
JDK-6310500 assert(SharedSkipVerify || obj->is_oop(),"sanity check")
- Closed