[code-reflection] RFR: Remove bytecode dialect [v8]

[Adam Sotona]

On Wed, 7 Feb 2024 19:18:18 GMT, Paul Sandoz <psandoz at openjdk.org> wrote:

> So there is something up with what's pushed onto the stack.

Yes, I wrongly assumed the stack can stay when transiting to a following block. That approach works for stack map generation or verification (where only types matter), however it is wrong here (where exact values are passed).

Fixed and added test.

Thanks.

-------------

PR Comment: https://git.openjdk.org/babylon/pull/16#issuecomment-1933692243