Fix small spec bugs regarding nullable reference type parameters

[jnm2]

While working on #1386, I found some spec bugs I could fix.

[jnm2]

This borrows the existing language which appears in https://github.com/dotnet/csharpstandard/blob/draft-v8/standard/types.md#8312-nullable-value-types:

Conversely, a non-nullable value type is any value type other than System.Nullable<T> and its shorthand T? (for any T), plus any type parameter that is constrained to be a non-nullable value type (that is, any type parameter with a value type constraint (§15.2.5)).

[jnm2]

Here is an example of a type parameter that is not unconstrained, and thus was not disallowed by the original statement, which is still not allowed to use ?:

#nullable enable

class C
{
    void M<T>(T? test) where T : System.IDisposable
    {
    }
}

Roslyn:

A nullable type parameter must be known to be a value type or non-nullable reference type unless language version '9.0' or greater is used. Consider changing the language version or adding a 'class', 'struct', or type constraint.

[jnm2]

Instances of T? do not exist when the type argument is a nullable reference type C? in C# 8, e.g.:

#nullable enable

class C
{
    void M<T>(T? test) where T : class?
    {
    }
}

Roslyn:

A nullable type parameter must be known to be a value type or non-nullable reference type unless language version '9.0' or greater is used. Consider changing the language version or adding a 'class', 'struct', or type constraint.

[BillWagner]

These all look good to me.

I'd like to have another set of eyes on them before merging, or have brief review in our next meeting.

/cc @jskeet

[jnm2]

Oops! Missed bringing this to the meeting since @jskeet didn't have a chance to look at it. I'll add the tag now.