RFR: 8279619: [JVMCI] improve EncodedSpeculationReason

[Doug Simon]

On Mon, 20 Feb 2023 18:09:25 GMT, Tom Rodriguez <never at openjdk.org> wrote:

>  I think most of the time the actual data is shorter than the size of a SHA-1 digest

In that case we just use the data [without hashing it](https://github.com/openjdk/jdk/blob/0bf3a53e01818aca5e365ee7275e567aad0273cc/src/jdk.internal.vm.ci/share/classes/jdk.vm.ci.hotspot/src/jdk/vm/ci/hotspot/HotSpotSpeculationEncoding.java#L205).

> our strategy for handing out ids protects us from this problem

Well it does now after we fixed a bug where the ids being handed out were not unique. In fact, that bug is why this issue exists. Surely it doesn't hurt to make this API more robust?

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

PR: https://git.openjdk.org/jdk/pull/12532