[API Proposal]: SVE: Correct multi register load methods

[a74nh]

#97831 approved the following:

namespace System.Runtime.Intrinsics.Arm

/// VectorT Summary
public abstract class Sve : AdvSimd /// Feature: FEAT_SVE  Category: loads
{
    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>) LoadVectorx2(Vector<T> mask, T* address); // LD2W or LD2D or LD2B or LD2H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>) LoadVectorx3(Vector<T> mask, T* address); // LD3W or LD3D or LD3B or LD3H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>, Vector<T>) LoadVectorx4(Vector<T> mask, T* address); // LD4W or LD4D or LD4B or LD4H
}

The following was implemented:

namespace System.Runtime.Intrinsics.Arm

/// VectorT Summary
public abstract class Sve : AdvSimd /// Feature: FEAT_SVE  Category: loads
{
    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>) LoadVector2xAndUnzip(Vector<T> mask, T* address); // LD2W or LD2D or LD2B or LD2H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>) LoadVector3xAndUnzip(Vector<T> mask, T* address); // LD3W or LD3D or LD3B or LD3H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>, Vector<T>) LoadVector4xAndUnzip(Vector<T> mask, T* address); // LD4W or LD4D or LD4B or LD4H
}

In additional there was no approval or implementation for the multiple structure LD1 instructions.

API Proposal

The following is proposed:

namespace System.Runtime.Intrinsics.Arm

/// VectorT Summary
public abstract class Sve : AdvSimd /// Feature: FEAT_SVE  Category: loads
{
    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>) LoadVector2x(Vector<T> mask, T* address); // LD1W or LD1D or LD1B or LD1H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>) LoadVector3x(Vector<T> mask, T* address); // LD1W or LD1D or LD1B or LD1H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>, Vector<T>) LoadVector4x(Vector<T> mask, T* address); // LD1W or LD1D or LD1B or LD1H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>) LoadVector2xAndUnzip(Vector<T> mask, T* address); // LD2W or LD2D or LD2B or LD2H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>) LoadVector3xAndUnzip(Vector<T> mask, T* address); // LD3W or LD3D or LD3B or LD3H

    /// T: float, double, sbyte, short, int, long, byte, ushort, uint, ulong
    public static unsafe (Vector<T>, Vector<T>, Vector<T>, Vector<T>) LoadVector4xAndUnzip(Vector<T> mask, T* address); // LD4W or LD4D or LD4B or LD4H
}

[dotnet-policy-service]

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

[tannergooding]

@a74nh, for AdvSimd we have both Load2xVector128 (LD1) and Load2xVector128AndUnzip (LD2)

My guess is we should have both Load2xVector (LD1W) and Load2xVectorAndUnzip (LD2W) -- plus variants for LD#D, LD#B, LD#Q, etc.

[dotnet-policy-service]

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

[kunalspathak]

For the naming differences, if we agree that the API names we implemented is correct, then we should just close this issue and open a new issue if there are additional variants missing.