The Java level range checks for StringLatin1::inflate() and StringUTF16::compress() are too weak. Offset and length need to be multiplied by two because they access a char value in a byte array:

--- a/src/java.base/share/classes/java/lang/StringLatin1.java Tue Apr 19 15:26:52 2016 -0400
+++ b/src/java.base/share/classes/java/lang/StringLatin1.java Thu Apr 28 10:30:23 2016 +0200
@@ -567,7 +567,7 @@
     @HotSpotIntrinsicCandidate
     public static void inflate(byte[] src, int srcOff, byte[] dst, int dstOff, int len) {
         // We need a range check here because 'putChar' has no checks
- checkBoundsOffCount(dstOff, len, dst.length);
+ checkBoundsOffCount(dstOff << 1, len << 1, dst.length);
         for (int i = 0; i < len; i++) {
             StringUTF16.putChar(dst, dstOff++, src[srcOff++] & 0xff);
         }

--- a/src/java.base/share/classes/java/lang/StringUTF16.java Tue Apr 19 15:26:52 2016 -0400
+++ b/src/java.base/share/classes/java/lang/StringUTF16.java Thu Apr 28 10:30:23 2016 +0200
@@ -162,7 +162,7 @@
     @HotSpotIntrinsicCandidate
     public static int compress(byte[] src, int srcOff, byte[] dst, int dstOff, int len) {
         // We need a range check here because 'getChar' has no checks
- checkBoundsOffCount(srcOff, len, src.length);
+ checkBoundsOffCount(srcOff << 1, len << 1, src.length);
         for (int i = 0; i < len; i++) {
             char c = getChar(src, srcOff);
             if (c > 0xFF) {