Revert "Improve ArgumentOutOfRange exception in Math.Clamp"

[jkotas]

Reverts #118824

[Copilot]

Pull Request Overview

This PR reverts changes from PR #118824 that improved ArgumentOutOfRange exceptions in Math.Clamp methods. The revert restores the original behavior of throwing a generic ArgumentException instead of the more specific ArgumentOutOfRangeException when the minimum value is greater than the maximum value in clamp operations.

Key changes:

• Reverts exception type from ArgumentOutOfRangeException back to ArgumentException for invalid min/max parameters
• Restores custom error message formatting instead of using the standard ArgumentOutOfRangeException helper
• Updates corresponding test expectations to match the reverted behavior

Reviewed Changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
src/libraries/System.Runtime/tests/System.Runtime.Extensions.Tests/System/Math.cs	Updates test expectations to verify ArgumentException is thrown instead of ArgumentOutOfRangeException
src/libraries/System.Runtime.Numerics/src/System/Numerics/BigInteger.cs	Reverts BigInteger.Clamp to use custom exception throwing with ArgumentException
src/libraries/System.Runtime.Numerics/src/Resources/Strings.resx	Adds back the custom error message format for min/max validation
src/libraries/System.Private.CoreLib/src/System/UInt128.cs	Reverts UInt128.Clamp to use Math.ThrowMinMaxException helper
src/libraries/System.Private.CoreLib/src/System/Single.cs	Reverts Single.Clamp and ClampNative to use Math.ThrowMinMaxException helper
src/libraries/System.Private.CoreLib/src/System/Math.cs	Reverts all Math.Clamp overloads and updates ThrowMinMaxException to throw ArgumentException
src/libraries/System.Private.CoreLib/src/System/Int128.cs	Reverts Int128.Clamp to use Math.ThrowMinMaxException helper
src/libraries/System.Private.CoreLib/src/System/Half.cs	Reverts Half.ClampNative to use Math.ThrowMinMaxException helper
src/libraries/System.Private.CoreLib/src/System/Double.cs	Reverts Double.Clamp and ClampNative to use Math.ThrowMinMaxException helper

[jkotas]

Context: #118824 (comment)

[dotnet-policy-service]

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

[xtqqczze]

If the implementation from the PR consistently threw when either min or max is NaN, that might be reasonable. However, it does not behave consistently, see sharplab.

[tannergooding]

The nuance around NaN handling and other considerations is part of why I think #119256 would be goodness.

It fixes a scenario that's often important for perf to be consistent across the stack (whether Scalar, Vector128, Tensors, etc), makes it consistent with what hardware typically does, and allows developers to easily define their own wrapper if they do want to throw/fail.

The biggest potential problem is it's a 4 year old API (added in .NET 6) that people may be depending on it throwing by this point.