SIG 11 Segmentation Violation in Java 1.4.2_03 on HP-UX 11.0

[Bojan_8]

We had a SIG 11 crash with core dump, twice within a week.
At crash time the eden heap space was at 100%, so I guess it has something to do with the garbage collector.
The executed function was follow_contents__15VirtualCallDataFv.

Somebody an idea?
We use a 1-year old Java version, maybe it's a known bug in 1.4.2_03 and fixed in latest version?

used versions:
HP-UX B.11.00 U 9000/800
Java HotSpot(TM) Server VM (build 1.4.2 1.4.2.03-040401-18:59-PA_RISC2.0 PA2.0 (aCC_AP), mixed mode)


Unexpected Signal : 11 occurred at PC=0xC69E10DB
Function=follow_contents__15VirtualCallDataFv
Library=/opt/java1.4.2_03/jre/lib/PA_RISC2.0/server/libjvm.sl


Dynamic libraries:
/opt/java1.4.2_03/bin/PA_RISC2.0/java
text:0x00001000-0x000098a4 data:0x0000a000-0x0000aa10

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/server/libjvm.sl
text:0xc6000000-0xc6c74000 data:0xbbe28000-0xbbfca000

/usr/lib/libpthread.1
text:0xc0c10000-0xc0c27000 data:0xbbde7000-0xbbdea000

/usr/lib/libm.2
text:0xc0060000-0xc008c000 data:0xbbdea000-0xbbdeb000

/usr/lib/librt.2
text:0xc030c000-0xc0310000 data:0xbbdeb000-0xbbdec000

/usr/lib/libcl.2
text:0xc0340000-0xc042b000 data:0xbbded000-0xbbdf9000

/usr/lib/libisamstub.1
text:0xc008f000-0xc0090000 data:0xbbdec000-0xbbded000

/usr/lib/libCsup.2
text:0xc0310000-0xc032b000 data:0xbbdf9000-0xbbdfc000

/usr/lib/libc.2
text:0xc0100000-0xc022f000 data:0xbbfcf000-0xbbfe6000

/usr/lib/libdld.2
text:0xc0003000-0xc0006000 data:0xbbfca000-0xbbfcb000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/native_threads/libhpi.sl
text:0xc0ad0000-0xc0ae0000 data:0xbbde1000-0xbbde3000

/usr/lib/libnss_files.1
text:0xc0008000-0xc000f000 data:0xbbdde000-0xbbddf000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libverify.sl
text:0xc0af0000-0xc0b00000 data:0xbbd5c000-0xbbd5d000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libjava.sl
text:0xc5cd0000-0xc5cfd000 data:0xbbd57000-0xbbd5c000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libzip.sl
text:0xc0b00000-0xc0b13000 data:0xbbd55000-0xbbd57000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libawt.sl
text:0xc9440000-0xc94e1000 data:0xbbc07000-0xbbc31000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libmlib_image.sl
text:0xc5f80000-0xc5ffa000 data:0xb9af3000-0xb9afe000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libmawt.sl
text:0xc9500000-0xc9571000 data:0xbbc01000-0xbbc05000

/usr/lib/libXm.4
text:0xc5400000-0xc5625000 data:0x3551b000-0x35555000

/usr/lib/X11R6/libXp.2
text:0xc1728000-0xc1730000 data:0xb9ab7000-0xb9ab8000

/usr/lib/X11R6/libXt.3
text:0xc5340000-0xc539d000 data:0xb9ac6000-0xb9ad1000

/usr/lib/X11R6/libSM.2
text:0xc2574000-0xc257d000 data:0xb9ac0000-0xb9ac2000

/usr/lib/X11R6/libICE.2
text:0xc5320000-0xc5335000 data:0xb9ac2000-0xb9ac6000

/usr/lib/X11R6/libXext.3
text:0xc5300000-0xc5317000 data:0xb9ad1000-0xb9ad3000

/usr/lib/X11R6/libXtst.2
text:0xc10f8000-0xc10fd000 data:0xbbc00000-0xbbc01000

/usr/lib/X11R6/libX11.3
text:0xc5200000-0xc52fe000 data:0xb9ad9000-0xb9af1000

/opt/java1.4.2_03/jre/lib/PA_RISC2.0/libnet.sl
text:0xc5630000-0xc5640000 data:0xb9ab6000-0xb9ab7000

/usr/lib/libnm.sl
text:0xc00ec000-0xc00f3000 data:0xb9ab5000-0xb9ab6000

/usr/lib/libnss_dns.1
text:0xc5625000-0xc5628000 data:0xb9ab4000-0xb9ab5000


Heap at VM Abort:
Heap
def new generation total 660224K, used 638683K [8f000000, b9aa0000, b9aa0000)
eden space 621440K, 100% used [8f000000, b4ee0000, b4ee0000)
from space 38784K, 44% used [b4ee0000, b5fb6e88, b74c0000)
to space 38784K, 0% used [b74c0000, b74c0000, b9aa0000)
tenured generation total 1398144K, used 764087K [39800000, 8ed60000, 8ed60000)
the space 1398144K, 54% used [39800000, 6822dd08, 6822de00, 8ed60000)
compacting perm gen total 27392K, used 27266K [35800000, 372c0000, 39800000)
the space 27392K, 99% used [35800000, 372a0830, 372a0a00, 372c0000)

Local Time = Tue Aug 23 11:52:41 2005
Elapsed Time = 78606
#
# HotSpot Virtual Machine Error : 11
# Please report this error to HP customer support.
#
# Java VM: Java HotSpot(TM) Server VM (1.4.2 1.4.2.03-040401-18:59-PA_RISC2.0 PA2.0 (aCC_AP) mixed mode)
#
# An error report file has been saved as hs_err_pid15166.log.
# Please refer to the file for further information.
#

[RAC_1]

For Java on HP, it is very good to keep your system updated on patches. Apart from that, be sure to check your initial, max heap sizes. Assign those values carefully.

Also, you may want to pursue it with HP.

[Bojan_8]

We are using a total heap size of 2GB with default values for other heap sub-spaces, concretely:
-Xms2048m -Xmx2048m -XX:+DisableLoopOptimization -XX:+ServerApp

The only improvement I see currently is to set the permanent size to 64m, as the default 16m is too small.

What do you mean by "pursue with HP"? If you mean logging a case at HP support, my manager just is searching the contract information...

[RAC_1]

Heap size of 2GB? Looks very bigh too me, but I am not Java expert, but still looks very high to me.

Pursue with HP, is take help from HP. Log a call with HP.