Fix nullable annotation for Validator.TryValidateValue and ValidateValue

[jeffhandley]

Fixes #91162

There was an erroneous comment in the Validator TryValidateValue and ValidateValue methods indicating that the value could not be null. This comment led to incorrect nullable annotations being applied. Propagation of that surfaced in the new Options source generator.

This removes the erroneous comment, updates the annotation to be more relaxed (non-breaking), and removes the ! from all call sites in the repo. The intent is to backport this fix to .NET 8 RC2.

[ghost]

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

Issue Details

---

Fixes #91162

There was an erroneous comment in the Validator TryValidateValue and ValidateValue methods indicating that the value could not be null. This comment led to incorrect nullable annotations being applied. Propagation of that surfaced in the new Options source generator.

This removes the erroneous comment, updates the annotation to be more relaxed (non-breaking), and removes the ! from all call sites in the repo. The intent is to backport this fix to .NET 8 RC2.

Author:	jeffhandley
Assignees:	jeffhandley
Labels:	area-System.ComponentModel.DataAnnotations
Milestone:	-

[tarekgh]

LGTM.

Are we planning to backport this one to 8.0 release? It makes sense to port it. I am preparing some other changes in the source gen which we'll need to port too. just FYI for now.

[jeffhandley]

/backport to release/8.0

[github-actions]

Started backporting to release/8.0: https://github.com/dotnet/runtime/actions/runs/6017311556

[stephentoub]

Do we have tests that validate everything compiles without warning/error with the options generator on .NET 6/7?

If it works, great, we should backport the change.

If not, I think we'll need to undo (or refine) the options generator change, so that it still uses ! when the parameters are non-nullable in the target reference assembly.

[jeffhandley]

Thanks for calling that out @stephentoub! I chatted with @tarekgh and I will push another PR that reverts the source generator references. He's going to consider if we could/should have the source generator be aware of this annotation difference.