-
Bug
-
Resolution: Not an Issue
-
P4
-
hs25, 8, 9
-
AMD x86 2191 MHz, 2 cores, 2G, Solaris / Solaris 10, i86pc
-
x86
-
solaris_10
The gc/arguments/TestUseCompressedOopsErgo.java jtreg test fails to launch a secondary VM when trying to verify heap bounds.
It tries to commit the global mark stack (32M) at that time, although there seems to be sufficient memory available 450M out of 2G).
Command line of the secondary VM: java -d64 -XX:+UseG1GC -Xmx34326183936 -XX:+PrintFlagsFinal -version.
Maybe setting a -Xms value in the test fixes this.
Stack trace:
V [libjvm.so+0x2805cc4] void VMError::report(outputStream*)+0x92c;; __1cHVMErrorGreport6MpnMoutputStream__v_+0x92c
V [libjvm.so+0x2807200] void VMError::report_and_die()+0x56c;; __1cHVMErrorOreport_and_die6M_v_+0x56c
V [libjvm.so+0xf81d2b] void report_vm_out_of_memory(const char*,int,unsigned long,VMErrorType,const char*)+0x497;; __1cXreport_vm_out_of_memory6FpkciLnLVMErrorType_1_v_+0x497
V [libjvm.so+0x21f3452] bool os::pd_commit_memory(char*,unsigned long,unsigned long,bool)+0x226;; __1cCosQpd_commit_memory6FpcLLb_b_+0x226
V [libjvm.so+0x21dcdb3] bool os::commit_memory(char*,unsigned long,unsigned long,bool)+0x23;; __1cCosNcommit_memory6FpcLLb_b_+0x23
V [libjvm.so+0x27ea01f] bool VirtualSpace::initialize(ReservedSpace,unsigned long)+0x34b;; __1cMVirtualSpaceKinitialize6MnNReservedSpace_L_b_+0x34b
V [libjvm.so+0xe82b2d] bool CMMarkStack::allocate(unsigned long)+0xa1;; __1cLCMMarkStackIallocate6ML_b_+0xa1
V [libjvm.so+0xe848eb] ConcurrentMark::ConcurrentMark #Nvariant 1(G1CollectedHeap*,ReservedSpace)+0xaab;; __1cOConcurrentMark2t6MpnPG1CollectedHeap_nNReservedSpace__v_+0xaab
V [libjvm.so+0x1152e8c] int G1CollectedHeap::initialize()+0x6f0;; __1cPG1CollectedHeapKinitialize6M_i_+0x6f0
V [libjvm.so+0x2708522] int Universe::initialize_heap()+0x47a;; __1cIUniversePinitialize_heap6F_i_+0x47a
V [libjvm.so+0x2706fe2] int universe_init()+0xea;; __1cNuniverse_init6F_i_+0xea
V [libjvm.so+0x1386c92] int init_globals()+0xa2;; __1cMinit_globals6F_i_+0xa2
V [libjvm.so+0x2687dce] int Threads::create_vm(JavaVMInitArgs*,bool*)+0x1d6;; __1cHThreadsJcreate_vm6FpnOJavaVMInitArgs_pb_i_+0x1d6
V [libjvm.so+0x1893bb0] JNI_CreateJavaVM+0x78;; JNI_CreateJavaVM+0x78
ILW: H (failed test), L (low, only occurence, M (run the test on a machine with more memory); P4 because apparently test configuration issue
It tries to commit the global mark stack (32M) at that time, although there seems to be sufficient memory available 450M out of 2G).
Command line of the secondary VM: java -d64 -XX:+UseG1GC -Xmx34326183936 -XX:+PrintFlagsFinal -version.
Maybe setting a -Xms value in the test fixes this.
Stack trace:
V [libjvm.so+0x2805cc4] void VMError::report(outputStream*)+0x92c;; __1cHVMErrorGreport6MpnMoutputStream__v_+0x92c
V [libjvm.so+0x2807200] void VMError::report_and_die()+0x56c;; __1cHVMErrorOreport_and_die6M_v_+0x56c
V [libjvm.so+0xf81d2b] void report_vm_out_of_memory(const char*,int,unsigned long,VMErrorType,const char*)+0x497;; __1cXreport_vm_out_of_memory6FpkciLnLVMErrorType_1_v_+0x497
V [libjvm.so+0x21f3452] bool os::pd_commit_memory(char*,unsigned long,unsigned long,bool)+0x226;; __1cCosQpd_commit_memory6FpcLLb_b_+0x226
V [libjvm.so+0x21dcdb3] bool os::commit_memory(char*,unsigned long,unsigned long,bool)+0x23;; __1cCosNcommit_memory6FpcLLb_b_+0x23
V [libjvm.so+0x27ea01f] bool VirtualSpace::initialize(ReservedSpace,unsigned long)+0x34b;; __1cMVirtualSpaceKinitialize6MnNReservedSpace_L_b_+0x34b
V [libjvm.so+0xe82b2d] bool CMMarkStack::allocate(unsigned long)+0xa1;; __1cLCMMarkStackIallocate6ML_b_+0xa1
V [libjvm.so+0xe848eb] ConcurrentMark::ConcurrentMark #Nvariant 1(G1CollectedHeap*,ReservedSpace)+0xaab;; __1cOConcurrentMark2t6MpnPG1CollectedHeap_nNReservedSpace__v_+0xaab
V [libjvm.so+0x1152e8c] int G1CollectedHeap::initialize()+0x6f0;; __1cPG1CollectedHeapKinitialize6M_i_+0x6f0
V [libjvm.so+0x2708522] int Universe::initialize_heap()+0x47a;; __1cIUniversePinitialize_heap6F_i_+0x47a
V [libjvm.so+0x2706fe2] int universe_init()+0xea;; __1cNuniverse_init6F_i_+0xea
V [libjvm.so+0x1386c92] int init_globals()+0xa2;; __1cMinit_globals6F_i_+0xa2
V [libjvm.so+0x2687dce] int Threads::create_vm(JavaVMInitArgs*,bool*)+0x1d6;; __1cHThreadsJcreate_vm6FpnOJavaVMInitArgs_pb_i_+0x1d6
V [libjvm.so+0x1893bb0] JNI_CreateJavaVM+0x78;; JNI_CreateJavaVM+0x78
ILW: H (failed test), L (low, only occurence, M (run the test on a machine with more memory); P4 because apparently test configuration issue
- relates to
-
JDK-8044132 Quarantine unstable/broken GC tests
- Resolved