JDK-6983135 : 50+ sajdi tests fail with UnalignedAddressException on WinXP with JDK7-B108 Client VM
  • Type: Bug
  • Component: hotspot
  • Sub-Component: svc
  • Affected Version: 7
  • Priority: P2
  • Status: Closed
  • Resolution: Duplicate
  • OS: windows_xp
  • CPU: x86
  • Submitted: 2010-09-08
  • Updated: 2012-02-01
  • Resolved: 2010-12-21
Related Reports
Duplicate :  
Relates :  
Description
During my baseline testing on JDK7-B108, 50+ VM/NSK sajdi tests failed
due to the following exception backtrace:

[2010-09-05T06:39:24.64] Caused by: sun.jvm.hotspot.debugger.UnalignedAddressException: Trying to read at address: 0x6dd28be5 with alignment: 4
[2010-09-05T06:39:24.64]        at sun.jvm.hotspot.debugger.windbg.WindbgDebuggerLocal$1.checkAlignment(WindbgDebuggerLocal.java:104)
[2010-09-05T06:39:24.64]        at sun.jvm.hotspot.debugger.DebuggerBase.readCInteger(DebuggerBase.java:349)
[2010-09-05T06:39:24.64]        at sun.jvm.hotspot.debugger.DebuggerBase.readAddressValue(DebuggerBase.java:454)
[2010-09-05T06:39:24.64]        at sun.jvm.hotspot.debugger.windbg.WindbgDebuggerLocal.readAddress(WindbgDebuggerLocal.java:316)
[2010-09-05T06:39:24.64]        at sun.jvm.hotspot.debugger.windbg.WindbgAddress.getAddressAt(WindbgAddress.java:72)
[2010-09-05T06:39:25.47]        at sun.jvm.hotspot.HotSpotTypeDataBase.readVMTypes(HotSpotTypeDataBase.java:125)
[2010-09-05T06:39:25.47]        at sun.jvm.hotspot.HotSpotTypeDataBase.<init>(HotSpotTypeDataBase.java:85)
[2010-09-05T06:39:25.47]        at sun.jvm.hotspot.HotSpotAgent.setupVM(HotSpotAgent.java:388)
[2010-09-05T06:39:25.47]        at sun.jvm.hotspot.HotSpotAgent.go(HotSpotAgent.java:315)
[2010-09-05T06:39:25.47]        at sun.jvm.hotspot.HotSpotAgent.attach(HotSpotAgent.java:158)
[2010-09-05T06:39:25.47]        at sun.jvm.hotspot.jdi.VirtualMachineImpl.createVirtualMachineForPID(VirtualMachineImpl.java:222)
[2010-09-05T06:39:25.47]        ... 11 more
[2010-09-05T06:39:25.47] script> Debugger finished with exit status: 1
[2010-09-05T06:39:25.47] script> # ERROR: Debugger class returned fail status: 1
[2010-09-05T06:39:25.47] script>
[2010-09-05T06:39:25.47] script> Kill debuggee VM process for pid: 6340
[2010-09-05T06:39:25.47] script> Kill process: 6340
[2010-09-05T06:39:25.47] $ C:/WINDOWS/system32/taskkill /F /PID 6340
[2010-09-05T06:39:25.47] SUCCESS: The process with PID 6340 has been terminated.
[2010-09-05T06:39:25.47] script>
[2010-09-05T06:39:25.47] script> Test exit status: 97


Results dir: vm-sajdi-prod-client-prod-comp.windows-i586
    executed: 93  passed: 18  ignored: 0  FAILED: 75
    UNEXPECTED: 52
    time: 26 minute(s)

Here is the list of tests:

nsk/sajdi/ArrayReference/getValue/getvalue001
nsk/sajdi/ArrayReference/getValues/getvalues001
nsk/sajdi/ArrayReference/getValues_ii/getvalues_ii001
nsk/sajdi/ClassLoaderReference/visibleClasses/vsbclasses001
nsk/sajdi/InterfaceType/superinterfaces/superintf001
nsk/sajdi/Method/allLineLocations/linelocs001
nsk/sajdi/Method/allLineLocations_ss/linelocs_ss001
nsk/sajdi/Method/bytecodes/bytecodes001
nsk/sajdi/Method/location/location001
nsk/sajdi/Method/locationsOfLine_i/locofline_i001
nsk/sajdi/Method/locationsOfLine_ssi/locofline_ssi001
nsk/sajdi/ObjectReference/getValues/getvalues001
nsk/sajdi/ObjectReference/owningThread/ownthrd002
nsk/sajdi/ReferenceType/allLineLocations/linelocs001
nsk/sajdi/ReferenceType/allLineLocations_ss/linelocs_ss001
nsk/sajdi/ReferenceType/fieldByName/fldbyname001
nsk/sajdi/ReferenceType/fields/fields001
nsk/sajdi/ReferenceType/getValues/getvalues001
nsk/sajdi/ReferenceType/locationsOfLine_i/locofline_i001
nsk/sajdi/ReferenceType/locationsOfLine_ssi/locofline_ssi001
nsk/sajdi/ReferenceType/methods/methods001
nsk/sajdi/ReferenceType/methodsByName/methbyname001
nsk/sajdi/ReferenceType/sourceDebugExtension/srcdebugx001
nsk/sajdi/SACoreAttachingConnector/attach/attach001
nsk/sajdi/SACoreAttachingConnector/attach/attach002
nsk/sajdi/SAPIDAttachingConnector/attach/attach001
nsk/sajdi/SAPIDAttachingConnector/attach/attach002
nsk/sajdi/ThreadGroupReference/threadGroups/thrdgrps001
nsk/sajdi/ThreadReference/ownedMonitors/ownmon002
nsk/sajdi/ThreadReference/status/status001
nsk/sajdi/VMCannotBeModifiedException/thrown/thrown001
nsk/sajdi/VMCannotBeModifiedException/thrown/thrown002
nsk/sajdi/VirtualMachine/allClasses/allclasses001
nsk/sajdi/VirtualMachine/allThreads/allthreads001
nsk/sajdi/VirtualMachine/canBeModified/canbemodif001
nsk/sajdi/VirtualMachine/capabilities/caps001
nsk/sajdi/VirtualMachine/capabilities/caps002
nsk/sajdi/VirtualMachine/classesByName/clsbyname001
nsk/sajdi/VirtualMachine/eventQueue/eventqueue001
nsk/sajdi/VirtualMachine/eventRequestManager/reqmgr001
nsk/sajdi/VirtualMachine/exit/exit001
nsk/sajdi/VirtualMachine/getDefaultStratum/getdfltstrat001
nsk/sajdi/VirtualMachine/mirrorOf/mirrorof001
nsk/sajdi/VirtualMachine/mirrorOf_s/mirrorof_s001
nsk/sajdi/VirtualMachine/process/process001
nsk/sajdi/VirtualMachine/redefineClasses/redefine001
nsk/sajdi/VirtualMachine/resume/resume001
nsk/sajdi/VirtualMachine/setDebugTraceMode/tracemode001
nsk/sajdi/VirtualMachine/setDefaultStratum/setdfltstrat001
nsk/sajdi/VirtualMachine/suspend/suspend001
nsk/sajdi/VirtualMachine/topLevelThreadGroups/topthrgroups001
nsk/sajdi/VirtualMachine/virtualMachine/virtmachine001

Note that the above are only the "unexpected failures". I suspect
that other tests also fail in this way, but are masked by other
entries on the fail_list.


I believe the following config fails the same test list:

Results dir: vm-sajdi-prod-client-prod-mixed.windows-i586
    executed: 93  passed: 18  ignored: 0  FAILED: 75
    UNEXPECTED: 52
    time: 27 minute(s)

Interestingly enough, the following two Client VM configs do not
have the same failures:

Results dir: vm-sajdi-prod-client-fast-comp.windows-i586
    executed: 93  passed: 89  ignored: 0  failed: 4
    time: 17 minute(s)
Results dir: vm-sajdi-prod-client-fast-mixed.windows-i586
    executed: 93  passed: 89  ignored: 0  failed: 4
    time: 14 minute(s)


The Server VM fails a similar list of tests in a different way.
I'll file a different bug for those failures.
The Server VM bug is:

6983140 3/3 50+ sajdi tests fail with Windbg Error: ReadVirtual
            failed on WinXP with JDK7-B108 Server VM
Here are the results for a rerun with JDK7-B107 Client VM bits:

Results dir: vm-sajdi-prod-client_b107-prod-comp.windows-i586
    executed: 93  passed: 89  ignored: 0  failed: 4
    time: 20 minute(s)
Results dir: vm-sajdi-prod-client_b107-prod-mixed.windows-i586
    executed: 93  passed: 89  ignored: 0  failed: 4
    time: 22 minute(s)

Summary of Test Results (2 result dirs)
=========================================
    all executed: 186  all passed: 178  all ignored: 0  all failed: 8
    time: 0 hour(s) 42 minute(s)
Here are the tests that fail with 'UnalignedAddressException'
on WinXP with JDK7-B108 Client VM that are obscured by other
bugs on the fail_list:

nsk/sajdi/ArrayType/componentType/comptype001
nsk/sajdi/ClassType/interfaces/interfaces001
nsk/sajdi/Method/arguments/arguments001
nsk/sajdi/Method/variables/variables001
nsk/sajdi/PathSearchingVirtualMachine/classPath/classpath001
nsk/sajdi/PrimitiveType/name/name001
nsk/sajdi/PrimitiveValue/value/value001
nsk/sajdi/ReferenceType/allFields/allfields001
nsk/sajdi/ReferenceType/allMethods/allmethods001
nsk/sajdi/ReferenceType/nestedTypes/nestedtypes001
nsk/sajdi/ReferenceType/nestedTypes/nestedtypes002
nsk/sajdi/ReferenceType/visibleFields/vsbfields001
nsk/sajdi/ReferenceType/visibleMethods/vsbmethods001
nsk/sajdi/SADebugServerAttachingConnector/attach/attach001
nsk/sajdi/SADebugServerAttachingConnector/attach/attach002
nsk/sajdi/SADebugServerAttachingConnector/attach/attach011
nsk/sajdi/SADebugServerAttachingConnector/attach/attach012
nsk/sajdi/ThreadReference/frames/frames001
nsk/sajdi/ThreadReference/status/status002
nsk/sajdi/jdb/options/connect/connect001
nsk/sajdi/jdb/options/connect/connect002
nsk/sajdi/jdb/options/connect/connect003
nsk/sajdi/jdb/options/connect/connect004

Comments
EVALUATION This bug describes a variation of the following bug: 6987812 2/3 SAJDI: "gHotSpotVMTypes was not initialized properly in the remote process" With the fix for 6987812 in place, this failure mode no longer occurs so I'm closing this bug as a duplicate of 6987812.
21-12-2010

EVALUATION A lot of tests fail
15-12-2010

EVALUATION There are a total of 149 tests that fail with UnalignedAddressException in my JDK7-B108 Win32 baseline testing: $ grep UnalignedAddressException vm-sajdi-prod-client-*/dcubed*/*/*.log | wc -l 149 However, in my JDK7-B109 Win32 baseline testing, there are none: $ grep UnalignedAddressException vm-sajdi-prod-client-*/dcubed*/*/*.log | wc -l 0 However, there are 148 tests that fail in the mode described by 6987812: $ grep 'gHotSpotVMTypes' vm-sajdi-prod-client-*/dcubed*/*/*.log | wc -l 148
28-09-2010