On Wed, 15 Jun 2022 22:27:27 GMT, Valerie Peng <valer...@openjdk.org> wrote:
>> Seems so. BTW, I was using a test-driven programming model and have not >> observed a leak here. I'll see if I can trigger one. > > Sounds good. Rest of changes look fine. I cannot reproduce this leak, but I found more by mixing the calls on byte array and `ByteBuffer`, and data length of whole block and half block. I'll study more and maybe some sort of fuzzy testing is needed. ------------- PR: https://git.openjdk.org/jdk/pull/9158
