Attached program sometimes causes crash on linux-i586. I was not able to reproduce it on other platforms. 
It is best reproduced with
java -XX:OnStackReplacePercentage=100 -XX:+UseConcMarkSweepGC -XX:+PrintCompilation -XX:CompileThreshold=100 gc.jit.GCDeopt.GCDeopt
Jstack is not able to get native stack trace from core file. Stack trace obtained using gdb follows.

#0  0xffffe410 in __kernel_vsyscall ()
#1  0xb7e700dd in raise () from /lib/tls/libc.so.6
#2  0xb7e716f3 in abort () from /lib/tls/libc.so.6
#3  0x062d0dbb in os::abort () from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
#4  0x06371391 in VMError::report_and_die () from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
#5  0x062d4cb0 in JVM_handle_linux_signal () from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
#6  0x062d2908 in signalHandler () from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
#7  <signal handler called>
#8  0x061cb2f0 in instanceKlass::lookup_osr_nmethod () from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
#9  0x061dea4b in InterpreterRuntime::frequency_counter_overflow ()
   from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
#10 0xb5d40d45 in ?? ()
#11 0x0805d900 in ?? ()
#12 0xb1bf0b88 in ?? ()
#13 0xb5d40d1a in ?? ()
#14 0xb7e090c8 in ?? ()
#15 0xb1bf0b76 in ?? ()
#16 0xb7e09100 in ?? ()
#17 0xb1bf2088 in ?? ()
#18 0x00000000 in ?? ()
#19 0xb1bf0be8 in ?? ()
#20 0x00000000 in ?? ()
#21 0xb7e09100 in ?? ()
#22 0xb7e09124 in ?? ()
#23 0xb5d34e9d in ?? ()
#24 0x00000001 in ?? ()
#25 0xade897f0 in ?? ()
#26 0xade897e0 in ?? ()
#27 0xade893c8 in ?? ()
#28 0xade89280 in ?? ()
#29 0xb7e09104 in ?? ()
#30 0xb1bf0ca6 in ?? ()
#31 0xb7e09134 in ?? ()
#32 0xb1bf2088 in ?? ()
#33 0x00000000 in ?? ()
#34 0xb1bf0cc0 in ?? ()
#35 0xb7e09100 in ?? ()
#36 0xb7e09134 in ?? ()
#37 0xb7e09158 in ?? ()
#38 0xb5d34e9d in ?? ()
#39 0x00000001 in ?? ()
#40 0xade89290 in ?? ()
#41 0xade89280 in ?? ()
#42 0xb7e09138 in ?? ()
---Type <return> to continue, or q <return> to quit---
#43 0xb1bf0d47 in ?? ()
#44 0xb7e09160 in ?? ()
#45 0xb1bf2088 in ?? ()
#46 0x00000000 in ?? ()
#47 0xb1bf0d50 in ?? ()
#48 0xb7e09134 in ?? ()
#49 0xb7e09160 in ?? ()
#50 0xb7e09178 in ?? ()
#51 0xb5d32207 in ?? ()
#52 0xade89270 in ?? ()
#53 0xb5b921e8 in ?? ()
#54 0x00001f80 in ?? ()
#55 0xb5d321a7 in ?? ()
#56 0x0000000a in ?? ()
#57 0x0805d900 in ?? ()
#58 0xb7e09228 in ?? ()
#59 0x061ea47d in JavaCalls::call_helper () from /net/vmsqe-amd-01/export3/vmsqe/jdk-builds/JDK6.0b60/linux-i586/jre/lib/i386/client/libjvm.so
Previous frame inner to this frame (corrupt stack?)

I am filing this under garbage collection category, because it is not reproducible without thread that does System.gc() in a loop.
Synopsys changed from:

Hotspot crash when concurrently doing GC and OSR

to:

 Crash in OSR / GC test

to avoid the ambiguous connotations of "concurrent" qualifier in
original synopsis.