[NativeAOT] consider using casting cache in GVMLookupForSlot scenario.

[VSadov]

It looks like GVMLookupForSlot has the same properties as casting - what is cached is relatively fast, so the cache needs to be faster, on the other hand it can have certain tradeoffs like simple/lossy resizing and expiration policies, while providing faster lookups and lock-free add.

It would make sense to try using Casting cache for GVMLookupForSlot.

Tagging subscribers to this area: @agocke, @MichalStrehovsky, @jkotas
See info in area-owners.md if you want to be subscribed.

Issue Details

---

It looks like GVMLookupForSlot has the same properties as casting - what is cached is relatively fast, so the cache needs to be faster, on the other can have certain tradeoffs like simple/lossy resizing and expiration policies.

It would make sense to try using Casting cache for GVMLookupForSlot.

Author:	VSadov
Assignees:	-
Labels:	area-NativeAOT-coreclr
Milestone:	-