JDK-4345058 : sun4m: JVM aborts immediately upon startup when "-client -Xcomp" switch is used.
  • Type: Bug
  • Component: hotspot
  • Sub-Component: compiler
  • Affected Version: 1.3.0
  • Priority: P1
  • Status: Resolved
  • Resolution: Fixed
  • OS: solaris_7
  • CPU: sparc
  • Submitted: 2000-06-13
  • Updated: 2000-07-24
  • Resolved: 2000-07-24
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
Other
1.3.0 sol-beta2Fixed
Related Reports
Relates :  
Description
Hostname: jtg-s206
Hostid: 727d45b9
Release: 5.7
Kernel architecture: sun4m
Application architecture: sparc
Hardware provider: Sun_Microsystems
Domain: SCTE.Eng.Sun.COM
Kernel version: SunOS 5.7 Generic 106541-11 March 2000

jtg-s206:[195]% psrinfo -v
Status of processor 0 as of: 06/12/00 18:59:18
  Processor has been on-line since 06/09/00 15:38:45.
  The sparc processor operates at 50 MHz,
        and has a sparc floating point processor.
Status of processor 2 as of: 06/12/00 18:59:18
  Processor has been on-line since 06/09/00 15:38:49.
  The sparc processor operates at 50 MHz,
        and has a sparc floating point processor.

Since JVM immediately aborts upon starting up, almost all Y2K tests and a great number of
JCK api/lang tests are failing consequently.

This problem is limited to "-client -Xcomp" ONLY, other switches or switch combinations 
will work fine.

To see a list of these failing testcases which involve with Y2K, JCK lang/api testsuites, point
browser to this URL:

  http://jtg-g1.eng/JTG_MASTER_ARCH/JDK_test/1.3.0-b18/sparc/SPARCstation-10/2/5.7/index.html
 

Run the following script to reproduce this problem:
(This script shows one case of the many y2k failures.)
==========================================================================
#!/bin/ksh

SWITCH=${1+$@}
TESTBASE=/net/ultraowl.eng/export/ultraowl1/javatb/testbase
JCK=${TESTBASE}/src/JCK-runtime-13fcs
PATH=/usr/java1.2/bin:/bin
CLASSPATH=.

export CLASSPATH PATH

java ${SWITCH} -version
echo
if [ ! -f y2k_1.java ]; then
  cp -p ${TESTBASE}/src/misc/y2k/y2k_1/y2k_1.java .
fi

echo 'Compiling ...'
javac ${1} -nowarn -d . *.java

echo 'Running ...'
java ${SWITCH} -Djava.security.policy=${JCK}/lib/jck.policy -verify y2k_1 

rc=$?
echo $rc ; return $rc

==========================================================================

jtg-s206:[222]% y2k_1.ksh -client -Xcomp
java version "1.3.0"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.0-b18)
Java HotSpot(TM) Client VM (build 1.3.0-b18, compiled mode)

Compiling ...
Running ...
#
# HotSpot Virtual Machine Error, Unexpected Signal 11
# Please report this error at
# http://java.sun.com/cgi-bin/bugreport.cgi
#
# Error ID: 4F533F534F4C415249530E4350500787 01
#
# Problematic Thread: prio=5 tid=0x9dbb0 nid=0xa runnable 
#
y2k_1.ksh[21]: 25095 Abort
134


vincent.lee@eng 2000-06-12

Provided more information after rerun the test case with java_g debugger.
#
# HotSpot Virtual Machine Error, assertion failure
# Please report this error at
# http://java.sun.com/cgi-bin/bugreport.cgi
#
# assert(end >= _instsStart && end <= _instsOverflow, "CodeBuffer overflow")
#
# Error ID: /export/sail1/jdk1.3_b18/hotsparc-1.3-b18/src/share/vm/asm/codeBuffer.hpp, 166 [ Patched ]
#
# Problematic Thread: prio=5 tid=0xecd50 nid=0xa runnable 
#
Dumping core....

david.kiang@Eng 2000-06-13

Comments
CONVERTED DATA BugTraq+ Release Management Values COMMIT TO FIX: kest-sol-beta2 FIXED IN: kest-sol-beta2 INTEGRATED IN: kest-sol-beta2
14-06-2004

EVALUATION The bug happens due to the fact the we run out of code buffer space during code emition. mohammad.gharahgouzloo@Eng 2000-06-14 added check in c1_CodeEmitter_sparc.cpp and bail out if there is not enough room in the buffer. The test was added to both slow case stub and call stub code generation. fix approved and OK for putback. mohammad.gharahgouzloo@Eng 2000-06-15
15-06-2000