Devin Papineau

The JIT changes LGTM, but we may also need to bump [`MINOR_NUMBER`](https://github.com/eclipse-openj9/openj9/blob/fbfa619e3c33b832bf600d1033ade3dcb37139db/runtime/compiler/net/CommunicationStream.hpp#L100). I'm thinking of a scenario where the client has continuation classes, but the server doesn't know about them,...

Commenting because I'd want your opinion before changing anything based on most of the following. ---- > the "Field offset" for say `Entity.velocity` in the Extended Class Chain's field data...

> Right, the idea behind that is what I was going for when I wrote "offset of the flattened fields in the field data refers to offsets in a class'...

> With checkpoint/restore, we're just assuming that portable code means it'll work on all machines. If it doesn't, isn't that the problem? The point of generating portable code is for...

The problem here is caused by out-of-bounds writes to the array `ax$0`. We've written past the end of it, and furthermore, it happens to have been stack-allocated by escape analysis,...

This depends on eclipse/omr#6712, and a coordinated merge will be needed

I haven't had a chance to dig in to this. Moved to 0.40.

I suspect that this is caused by a method handle table entry initialization race condition together with the JIT's handling of known objects, and that the apparent connection to AOT...

I'm reviewing now

I'll review soon