[wasm] Add PackedSimd saturating integer arith.

[radekdoulik]

No description provided.

[ghost]

Note regarding the new-api-needs-documentation label:

This serves as a reminder for when your PR is modifying a ref *.cs file and adding/modifying public APIs, please make sure the API implementation in the src *.cs file is documented with triple slash comments, so the PR reviewers can sign off that change.

[ghost]

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

Issue Details

---

null

Author:	radekdoulik
Assignees:	radekdoulik
Labels:	area-System.Runtime.Intrinsics, new-api-needs-documentation
Milestone:	-

[radekdoulik]

A test method:

    static void TestSatA()
    {
        Vector128<int> v1int = Vector128.Create(Random.Shared.Next(), Random.Shared.Next(), Random.Shared.Next(), Random.Shared.Next());
        Vector128<int> v2int = Vector128.Create(Random.Shared.Next(), Random.Shared.Next(), Random.Shared.Next(), Random.Shared.Next());

        Vector128<byte> v1byte = Vector128.Create((byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next());
        Vector128<byte> v2byte = Vector128.Create((byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next(), (byte)Random.Shared.Next());

        Vector128<short> v1short = Vector128.Create((short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next());
        Vector128<short> v2short = Vector128.Create((short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next(), (short)Random.Shared.Next());

        Console.WriteLine($"v1: {v1int}");
        Console.WriteLine($"v2: {v2int}");
        Console.WriteLine($"add sat: {PackedSimd.AddSaturate(v1byte, v2byte)}");
        Console.WriteLine($"sub sat: {PackedSimd.SubtractSaturate(v1byte, v2byte)}");
        Console.WriteLine($"mul round sat: {PackedSimd.MultiplyRoundedSaturateQ15(v1short, v2short)}");
        Console.WriteLine($"min: {PackedSimd.Min(v1int, v2int)}");
        Console.WriteLine($"max: {PackedSimd.Max(v1int, v2int)}");
        Console.WriteLine($"avg round sat: {PackedSimd.AverageRounded(v1byte, v2byte)}");
        Console.WriteLine($"abs: {PackedSimd.Abs(v1int)}");
    }

output:

  v1: <1050742669, 1570759397, 963964663, 1067018614>
  v2: <1111389472, 376157844, 484061419, 917414378>
  add sat: <255, 188, 255, 255, 84, 255, 146, 193, 255, 255, 220, 255, 255, 217, 196, 184>
  sub sat: <0, 0, 31, 0, 16, 26, 54, 0, 0, 0, 136, 100, 12, 0, 0, 14>
  mul round sat: <18625, -17392, -14101, 6577, 25508, -971, 10840, -8814>
  min: <1111389472, 1570759397, 963964663, 1067018614>
  max: <1050742669, 376157844, 484061419, 917414378>
  avg round sat: <138, 94, 239, 134, 42, 181, 73, 97, 131, 135, 110, 163, 145, 109, 98, 92>
  abs: <1050742669, 1570759397, 963964663, 1067018614>

and instructions in the emitted code:

> wa-info -d -f TestSat /mnt/d/git/wa-local/src/mono/sample/wasm/console-node/bin/Release/AppBundle/dotnet.wasm|grep SIMD|egrep 'sat|abs|avg|min|max'
     i8x16.add.sat.u    [SIMD]
     i8x16.sub.sat.u    [SIMD]
     i16x8.q15mulr.sat.s    [SIMD]
     i32x4.max.s    [SIMD]
     i32x4.min.s    [SIMD]
     i8x16.avgr.u    [SIMD]
 i32x4.abs    [SIMD]

[vargaz]

The mono changes look ok.

[lewing]

These are implementing the already approved packed simd