Skip to content

Update BitArray.CopyTo to use cross-platform intrinsics #120627

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Oct 15, 2025
Merged

Update BitArray.CopyTo to use cross-platform intrinsics #120627

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e457e53
Initial plan
Copilot Oct 11, 2025
51e0e81
Update BitArray.CopyTo to use cross-platform intrinsics
Copilot Oct 11, 2025
373b6b0
Update Vector128 paths to use cross-platform intrinsics and consolida…
Copilot Oct 12, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
101 changes: 19 additions & 82 deletions src/libraries/System.Private.CoreLib/src/System/Collections/BitArray.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -738,7 +738,7 @@ public unsafe void CopyTo(Array array, int index)
Vector128<byte> lowerShuffleMask_CopyToBoolArray = Vector128.Create(0, 0x01010101_01010101).AsByte();
Vector128<byte> upperShuffleMask_CopyToBoolArray = Vector128.Create(0x02020202_02020202, 0x03030303_03030303).AsByte();

if (Avx512BW.IsSupported && (uint)_bitLength >= Vector512<byte>.Count)
if (Vector512.IsHardwareAccelerated && (uint)_bitLength >= Vector512<byte>.Count)
{
Vector256<byte> upperShuffleMask_CopyToBoolArray256 = Vector256.Create(0x04040404_04040404, 0x05050505_05050505,
0x06060606_06060606, 0x07070707_07070707).AsByte();
Expand All @@ -753,17 +753,17 @@ public unsafe void CopyTo(Array array, int index)
{
ulong bits = (ulong)(uint)in32Span[(int)(i / (uint)BitsPerInt32)] + ((ulong)in32Span[(int)(i / (uint)BitsPerInt32) + 1] << BitsPerInt32);
Vector512<ulong> scalar = Vector512.Create(bits);
Vector512<byte> shuffled = Avx512BW.Shuffle(scalar.AsByte(), shuffleMask);
Vector512<byte> extracted = Avx512F.And(shuffled, bitMask);
Vector512<byte> shuffled = Vector512.Shuffle(scalar.AsByte(), shuffleMask);
Vector512<byte> extracted = shuffled & bitMask;

// The extracted bits can be anywhere between 0 and 255, so we normalise the value to either 0 or 1
// to ensure compatibility with "C# bool" (0 for false, 1 for true, rest undefined)
Vector512<byte> normalized = Avx512BW.Min(extracted, ones);
Avx512F.Store((byte*)destination + i, normalized);
Vector512<byte> normalized = Vector512.Min(extracted, ones);
normalized.Store((byte*)destination + i);
}
}
}
else if (Avx2.IsSupported && (uint)_bitLength >= Vector256<byte>.Count)
else if (Vector256.IsHardwareAccelerated && (uint)_bitLength >= Vector256<byte>.Count)
{
Vector256<byte> shuffleMask = Vector256.Create(lowerShuffleMask_CopyToBoolArray, upperShuffleMask_CopyToBoolArray);
Vector256<byte> bitMask = Vector256.Create(0x80402010_08040201).AsByte();
Expand All @@ -775,17 +775,17 @@ public unsafe void CopyTo(Array array, int index)
{
int bits = in32Span[(int)(i / (uint)BitsPerInt32)];
Vector256<int> scalar = Vector256.Create(bits);
Vector256<byte> shuffled = Avx2.Shuffle(scalar.AsByte(), shuffleMask);
Vector256<byte> extracted = Avx2.And(shuffled, bitMask);
Vector256<byte> shuffled = Vector256.Shuffle(scalar.AsByte(), shuffleMask);
Vector256<byte> extracted = shuffled & bitMask;

// The extracted bits can be anywhere between 0 and 255, so we normalise the value to either 0 or 1
// to ensure compatibility with "C# bool" (0 for false, 1 for true, rest undefined)
Vector256<byte> normalized = Avx2.Min(extracted, ones);
Avx.Store((byte*)destination + i, normalized);
Vector256<byte> normalized = Vector256.Min(extracted, ones);
normalized.Store((byte*)destination + i);
}
}
}
else if (Ssse3.IsSupported && ((uint)_bitLength >= Vector128<byte>.Count * 2u))
else if (Vector128.IsHardwareAccelerated && ((uint)_bitLength >= Vector128<byte>.Count * 2u))
{
Vector128<byte> lowerShuffleMask = lowerShuffleMask_CopyToBoolArray;
Vector128<byte> upperShuffleMask = upperShuffleMask_CopyToBoolArray;
Expand All @@ -799,82 +799,19 @@ public unsafe void CopyTo(Array array, int index)
int bits = in32Span[(int)(i / (uint)BitsPerInt32)];
Vector128<int> scalar = Vector128.CreateScalarUnsafe(bits);

Vector128<byte> shuffledLower = Ssse3.Shuffle(scalar.AsByte(), lowerShuffleMask);
Vector128<byte> extractedLower = Sse2.And(shuffledLower, bitMask128);
Vector128<byte> normalizedLower = Sse2.Min(extractedLower, ones);
Sse2.Store((byte*)destination + i, normalizedLower);
Vector128<byte> shuffledLower = Vector128.Shuffle(scalar.AsByte(), lowerShuffleMask);
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's likely worth noting that since this is byte, there shouldn't be any considerations. However, Vector.Shuffle has to have deterministic behavior across platforms for "out of range" indices and so, for other T, can slightly pessimize to handle that scenario.

We're using a "vector constant" here, but its indirect and through a couple local copies. The JIT should still see it in such other scenarios and the cost is minimal, so shouldn't matter. But it's something to note anyways.

Vector128<byte> extractedLower = shuffledLower & bitMask128;
Vector128<byte> normalizedLower = Vector128.Min(extractedLower, ones);
normalizedLower.Store((byte*)destination + i);

Vector128<byte> shuffledHigher = Ssse3.Shuffle(scalar.AsByte(), upperShuffleMask);
Vector128<byte> extractedHigher = Sse2.And(shuffledHigher, bitMask128);
Vector128<byte> normalizedHigher = Sse2.Min(extractedHigher, ones);
Sse2.Store((byte*)destination + i + Vector128<byte>.Count, normalizedHigher);
Vector128<byte> shuffledHigher = Vector128.Shuffle(scalar.AsByte(), upperShuffleMask);
Vector128<byte> extractedHigher = shuffledHigher & bitMask128;
Vector128<byte> normalizedHigher = Vector128.Min(extractedHigher, ones);
normalizedHigher.Store((byte*)destination + i + Vector128<byte>.Count);
}
}
}
else if (PackedSimd.IsSupported && ((uint)_bitLength >= Vector128<byte>.Count * 2u))
{
Vector128<byte> lowerShuffleMask = lowerShuffleMask_CopyToBoolArray;
Vector128<byte> upperShuffleMask = upperShuffleMask_CopyToBoolArray;
Vector128<byte> ones = Vector128<byte>.One;
Vector128<byte> bitMask128 = Vector128.Create(0x80402010_08040201).AsByte();

fixed (bool* destination = &boolArray[index])
{
for (; (i + Vector128<byte>.Count * 2u) <= (uint)_bitLength; i += (uint)Vector128<byte>.Count * 2u)
{
int bits = in32Span[(int)(i / (uint)BitsPerInt32)];
Vector128<int> scalar = Vector128.CreateScalarUnsafe(bits);

Vector128<byte> shuffledLower = PackedSimd.Swizzle(scalar.AsByte(), lowerShuffleMask);
Vector128<byte> extractedLower = PackedSimd.And(shuffledLower, bitMask128);
Vector128<byte> normalizedLower = PackedSimd.Min(extractedLower, ones);
PackedSimd.Store((byte*)destination + i, normalizedLower);

Vector128<byte> shuffledHigher = PackedSimd.Swizzle(scalar.AsByte(), upperShuffleMask);
Vector128<byte> extractedHigher = PackedSimd.And(shuffledHigher, bitMask128);
Vector128<byte> normalizedHigher = PackedSimd.Min(extractedHigher, ones);
PackedSimd.Store((byte*)destination + i + Vector128<byte>.Count, normalizedHigher);
}
}
}
else if (AdvSimd.Arm64.IsSupported)
{
Vector128<byte> ones = Vector128<byte>.One;
Vector128<byte> bitMask128 = Vector128.Create(0x80402010_08040201).AsByte();

fixed (bool* destination = &boolArray[index])
{
for (; (i + Vector128<byte>.Count * 2u) <= (uint)_bitLength; i += (uint)Vector128<byte>.Count * 2u)
{
int bits = in32Span[(int)(i / (uint)BitsPerInt32)];

// Same logic as SSSE3 path, except we do not have Shuffle instruction.
// (TableVectorLookup could be an alternative - dotnet/runtime#1277)
// Instead we use chained ZIP1/2 instructions:
// (A0 is the byte containing LSB, A3 is the byte containing MSB)
// bits - A0 A1 A2 A3
// v1 = Vector128.Create - A0 A1 A2 A3 A0 A1 A2 A3 A0 A1 A2 A3 A0 A1 A2 A3
// v2 = ZipLow(v1, v1) - A0 A0 A1 A1 A2 A2 A3 A3 A0 A0 A1 A1 A2 A2 A3 A3
// v3 = ZipLow(v2, v2) - A0 A0 A0 A0 A1 A1 A1 A1 A2 A2 A2 A2 A3 A3 A3 A3
// shuffledLower = ZipLow(v3, v3) - A0 A0 A0 A0 A0 A0 A0 A0 A1 A1 A1 A1 A1 A1 A1 A1
// shuffledHigher = ZipHigh(v3, v3) - A2 A2 A2 A2 A2 A2 A2 A2 A3 A3 A3 A3 A3 A3 A3 A3

Vector128<byte> vector = Vector128.Create(bits).AsByte();
vector = AdvSimd.Arm64.ZipLow(vector, vector);
vector = AdvSimd.Arm64.ZipLow(vector, vector);

Vector128<byte> shuffledLower = AdvSimd.Arm64.ZipLow(vector, vector);
Vector128<byte> extractedLower = AdvSimd.And(shuffledLower, bitMask128);
Vector128<byte> normalizedLower = AdvSimd.Min(extractedLower, ones);

Vector128<byte> shuffledHigher = AdvSimd.Arm64.ZipHigh(vector, vector);
Vector128<byte> extractedHigher = AdvSimd.And(shuffledHigher, bitMask128);
Vector128<byte> normalizedHigher = AdvSimd.Min(extractedHigher, ones);

AdvSimd.Arm64.StorePair((byte*)destination + i, normalizedLower, normalizedHigher);
}
}
}

Remainder:
for (; i < (uint)_bitLength; i++)
Expand Down
Loading