AWT crashes inside CCombinedSegTable::In called from Java_sun_awt_windows_WDefaultFontCharset_canConvert

AWT crashes inside CCombinedSegTable::In called from Java_sun_awt_windows_WDefaultFontCharset_canConvert:

#
# A fatal error has been detected by the Java Runtime Environment:
#
# EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x0000000072005c9c, pid=12464, tid=9932
#
# JRE version: Java(TM) SE Runtime Environment (8.0_20-b16) (build 1.8.0_20-ea-b16)
# Java VM: Java HotSpot(TM) 64-Bit Server VM (25.20-b16 mixed mode windows-amd64 compressed oops)
# Problematic frame:
# C [awt.dll+0xa5c9c] CCombinedSegTable::In+0x14
#
# Failed to write core dump. Minidumps are not enabled by default on client versions of Windows
#
# If you would like to submit a bug report, please visit:
# http://bugreport.sun.com/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#
....
Stack: [0x0000000046b60000,0x0000000046c60000], sp=0x0000000046c5db10, free space=1014k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
C [awt.dll+0xa5c9c] CCombinedSegTable::In+0x14
C [awt.dll+0xa66bb] Java_sun_awt_windows_WDefaultFontCharset_canConvert+0xcf
C 0x00000000029940bf

Java frames: (J=compiled Java code, j=interpreted, Vv=VM code)
j sun.awt.windows.WDefaultFontCharset.canConvert(C)Z+0
j sun.awt.windows.WDefaultFontCharset.access$100(Lsun/awt/windows/WDefaultFontCharset;C)Z+2
j sun.awt.windows.WDefaultFontCharset$Encoder.canEncode(C)Z+5
j sun.awt.PlatformFont.<init>(Ljava/lang/String;I)V+160
j sun.awt.windows.WFontPeer.<init>(Ljava/lang/String;I)V+3
j sun.awt.windows.WToolkit.getFontPeer(Ljava/lang/String;I)Ljava/awt/peer/FontPeer;+59
j java.awt.Font.getPeer_NoClientCode()Ljava/awt/peer/FontPeer;+21
j java.awt.Font.getPeer()Ljava/awt/peer/FontPeer;+1
v ~StubRoutines::call_stub
j sun.awt.windows.WComponentPeer._setFont(Ljava/awt/Font;)V+0
j sun.awt.windows.WComponentPeer.setFont(Ljava/awt/Font;)V+7
j sun.awt.windows.WWindowPeer.initialize()V+44
j sun.awt.windows.WFramePeer.initialize()V+1
j sun.awt.windows.WComponentPeer.<init>(Ljava/awt/Component;)V+83
j sun.awt.windows.WCanvasPeer.<init>(Ljava/awt/Component;)V+2
j sun.awt.windows.WPanelPeer.<init>(Ljava/awt/Component;)V+2
j sun.awt.windows.WWindowPeer.<init>(Ljava/awt/Window;)V+2
j sun.awt.windows.WFramePeer.<init>(Ljava/awt/Frame;)V+2
j sun.awt.windows.WToolkit.createFrame(Ljava/awt/Frame;)Ljava/awt/peer/FramePeer;+5
j java.awt.Frame.addNotify()V+20
j java.awt.Window.show()V+8
j java.awt.Component.show(Z)V+5
j java.awt.Component.setVisible(Z)V+2
j java.awt.Window.setVisible(Z)V+2
j javasoft.sqe.tests.api.java.awt.Component.isValidTests.Component2008()Ljavasoft/sqe/harness/Status;+27
v ~StubRoutines::call_stub
j sun.reflect.NativeMethodAccessorImpl.invoke0(Ljava/lang/reflect/Method;Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+0
j sun.reflect.NativeMethodAccessorImpl.invoke(Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+100
j sun.reflect.DelegatingMethodAccessorImpl.invoke(Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+6
j java.lang.reflect.Method.invoke(Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+56
j javasoft.sqe.jck.lib.MultiTest.run([Ljava/lang/String;Ljava/io/PrintStream;Ljava/io/PrintStream;)Ljavasoft/sqe/harness/Status;+147
j sun.reflect.GeneratedMethodAccessor2.invoke(Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+56
j sun.reflect.DelegatingMethodAccessorImpl.invoke(Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+6
j java.lang.reflect.Method.invoke(Ljava/lang/Object;[Ljava/lang/Object;)Ljava/lang/Object;+56
j nsk.stress.share.StressTest$TestThread.runTest(I)V+250
j nsk.stress.share.StressTest$TestThread.run()V+27
v ~StubRoutines::call_stub

Issue could be reproduced with VM stress test that execute many different JCK tests concurrently. See comments for more details.

Issue could be reproduced starting from 8u20-b11.

I was able to reproduce issue w/ latest 8u20 and 9 builds.
I was not able to reproduce it w/ 8 FCS, so apparently it is a regression.

hs-err and compressed minidump attached to CR.