Investigate ways to make GC stress compatible with WaitForPendingFinalizers

[VSadov]

See a comment in #97334 (comment)
and discussion in: #100576

Currently it is way too easy to write a pattern that uses self-rearming finalizable object that would turn into never ending finalization queue under GC stress. Some such cases may be present in the runtime/libraries already.

As a result it is easy for a test that calls WaitForPendingFinalizers to hang under GC stress.
This applies not only to tests and GC test legs, but also to 3-rd party use of GC stress to validate their apps.

It would be nice to make GC stress tolerant to such scenarios. One proposed way is to disable GC stress for the duration of WaitForPendingFinalizers calls. (perhaps with a counter, for concurrent cases)

[dotnet-policy-service]

Tagging subscribers to this area: @dotnet/runtime-infrastructure
See info in area-owners.md if you want to be subscribed.

[dotnet-policy-service]

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.