Regression brought by optimization done with JDK-4926314

Supposedly after the fix for JDK-4926314 the following code

    new BufferedReader(new CharArrayReader(new char[]{1, 2, 3}))
                            .read(CharBuffer.allocate(0))

started to produce '-1' instead of the expected '0'

Specification for method Reader::read says:

"Returns: The number of characters added to the buffer, or -1 if this source of characters is at its end"