Backport reflection registration fix for JFR eventHandler fields

[jerboaa]

Backports: oracle@bdaf8c2

Fix reflection registration for JFR eventHandler fields
Closes #270

[zakkak]

FTR (and for linking the GH PRs) the corresponding upstream PR was oracle#3662

[zakkak]

I would prefer the whole PR (oracle#3662) backported, i.e., I would prefer to see 7496416 and e7476d8 backported as well. WDYT?

[jiekang]

I would prefer the whole PR (oracle#3662) backported, i.e., I would prefer to see 7496416 and e7476d8 backported as well. WDYT?

Yes, I'd like to see those as well. e7476d8 has a meaningful change.

[jerboaa]

Thanks for the reviews! I'll fix.

Sigh. Working with Graal VM is a bit of a mess. So you really need to be looking at each and every PR to get the right commits? Wow. I guess squashing of commits makes more sense in this model. Looking at the Git logs it doesn't seem to be the upstream style, though...

[zakkak]

Sigh. Working with Graal VM is a bit of a mess. So you really need to be looking at each and every PR to get the right commits? Wow. I guess squashing of commits makes more sense in this model. Looking at the Git logs it doesn't seem to be the upstream style, though...

Well, trade offs :)

Thanks for the backport Severin!