Skip to content

Vectorize TrimTransparentPixels in GifEncoderCore #2500

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
JimBobSquarePants merged 2 commits into from
Aug 9, 2023
Merged

Vectorize TrimTransparentPixels in GifEncoderCore #2500

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5416edb
Vectorize TrimTransparentPixels in GifEncoderCore
gfoidl Jul 27, 2023
c8f1f2c
Simplified check if there are any non-equal bytes
gfoidl Jul 27, 2023
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
133 changes: 126 additions & 7 deletions src/ImageSharp/Formats/Gif/GifEncoderCore.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -412,23 +412,142 @@ private static Buffer2DRegion<byte> TrimTransparentPixels(Buffer2D<byte> buffer,
int bottom = int.MaxValue;
int left = int.MaxValue;
int right = int.MinValue;

// Run through th buffer in a single pass. Use variables to track the min/max values.
int minY = -1;
bool isTransparentRow = true;

// Run through the buffer in a single pass. Use variables to track the min/max values.
for (int y = 0; y < buffer.Height; y++)
{
isTransparentRow = true;
Span<byte> rowSpan = buffer.DangerousGetRowSpan(y);
ref byte rowPtr = ref MemoryMarshal.GetReference(rowSpan);
nint rowLength = (nint)(uint)rowSpan.Length;
nint x = 0;

#if NET7_0_OR_GREATER
if (Vector128.IsHardwareAccelerated && rowLength >= Vector128<byte>.Count)
{
Vector256<byte> trimmableVec256 = Vector256.Create(trimmableIndex);

if (Vector256.IsHardwareAccelerated && rowLength >= Vector256<byte>.Count)
{
do
{
Vector256<byte> vec = Vector256.LoadUnsafe(ref rowPtr, (nuint)x);
Vector256<byte> notEquals = ~Vector256.Equals(vec, trimmableVec256);

if (notEquals != Vector256<byte>.Zero)
Copy link
Contributor Author

@gfoidl gfoidl Jul 27, 2023

Choose a reason for hiding this comment

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

At the moment I don't have any idea on how to make this branchless.
isTransparentRow could be tracked in a vector, but left and right not, as there's a mismatch of vector-types, namely byte and int.

A quite complicated approach would be to use VectorXYZ<byte> and track the left and right -- but just before these can overflow merge it back to the scalar left, right and start over. But I guess the book-keeping is more work, so I'm not sure if this is actually faster. For sure the code gets painful.

{
isTransparentRow = false;
uint mask = notEquals.ExtractMostSignificantBits();
nint start = x + (nint)uint.TrailingZeroCount(mask);
nint end = (nint)uint.LeadingZeroCount(mask);

// end is from the end, but we need the index from the beginning
end = x + Vector256<byte>.Count - 1 - end;

left = Math.Min(left, (int)start);
right = Math.Max(right, (int)end);
}

x += Vector256<byte>.Count;
}
while (x <= rowLength - Vector256<byte>.Count);
}

Vector128<byte> trimmableVec = Vector256.IsHardwareAccelerated
? trimmableVec256.GetLower()
: Vector128.Create(trimmableIndex);

while (x <= rowLength - Vector128<byte>.Count)
{
Vector128<byte> vec = Vector128.LoadUnsafe(ref rowPtr, (nuint)x);
Vector128<byte> notEquals = ~Vector128.Equals(vec, trimmableVec);

if (notEquals != Vector128<byte>.Zero)
{
isTransparentRow = false;
uint mask = notEquals.ExtractMostSignificantBits();
nint start = x + (nint)uint.TrailingZeroCount(mask);
nint end = (nint)uint.LeadingZeroCount(mask) - Vector128<byte>.Count;

// end is from the end, but we need the index from the beginning
end = x + Vector128<byte>.Count - 1 - end;

left = Math.Min(left, (int)start);
right = Math.Max(right, (int)end);
}

x += Vector128<byte>.Count;
}
}
#else
if (Sse41.IsSupported && rowLength >= Vector128<byte>.Count)
{
Vector256<byte> trimmableVec256 = Vector256.Create(trimmableIndex);

if (Avx2.IsSupported && rowLength >= Vector256<byte>.Count)
{
do
{
Vector256<byte> vec = Unsafe.ReadUnaligned<Vector256<byte>>(ref Unsafe.Add(ref rowPtr, x));
Vector256<byte> notEquals = Avx2.CompareEqual(vec, trimmableVec256);
notEquals = Avx2.Xor(notEquals, Vector256<byte>.AllBitsSet);

if (!Avx.TestZ(notEquals, notEquals))
{
isTransparentRow = false;
int mask = Avx2.MoveMask(notEquals);
nint start = x + (nint)(uint)BitOperations.TrailingZeroCount(mask);
nint end = (nint)(uint)BitOperations.LeadingZeroCount((uint)mask);

// end is from the end, but we need the index from the beginning
end = x + Vector256<byte>.Count - 1 - end;

left = Math.Min(left, (int)start);
right = Math.Max(right, (int)end);
}

x += Vector256<byte>.Count;
}
while (x <= rowLength - Vector256<byte>.Count);
}

Vector128<byte> trimmableVec = Sse41.IsSupported
? trimmableVec256.GetLower()
: Vector128.Create(trimmableIndex);

while (x <= rowLength - Vector128<byte>.Count)
{
Vector128<byte> vec = Unsafe.ReadUnaligned<Vector128<byte>>(ref Unsafe.Add(ref rowPtr, x));
Vector128<byte> notEquals = Sse2.CompareEqual(vec, trimmableVec);
notEquals = Sse2.Xor(notEquals, Vector128<byte>.AllBitsSet);

if (!Sse41.TestZ(notEquals, notEquals))
{
isTransparentRow = false;
int mask = Sse2.MoveMask(notEquals);
nint start = x + (nint)(uint)BitOperations.TrailingZeroCount(mask);
nint end = (nint)(uint)BitOperations.LeadingZeroCount((uint)mask) - Vector128<byte>.Count;

// TODO: It may be possible to optimize this inner loop using SIMD.
for (int x = 0; x < rowSpan.Length; x++)
// end is from the end, but we need the index from the beginning
end = x + Vector128<byte>.Count - 1 - end;

left = Math.Min(left, (int)start);
right = Math.Max(right, (int)end);
}

x += Vector128<byte>.Count;
}
}
#endif
for (; x < rowLength; ++x)
Copy link
Contributor Author

@gfoidl gfoidl Jul 27, 2023

Choose a reason for hiding this comment

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

The remainder could be handled vectorized too, by shifting the mask of the most significant bits around by the count of elements left in the final vector.
I tried this somewhere else, the cost for that book-keeping isn't negligible, so didn't do this here and now (maybe I'll try this later).

{
if (rowSpan[x] != trimmableIndex)
if (Unsafe.Add(ref rowPtr, x) != trimmableIndex)
{
isTransparentRow = false;
left = Math.Min(left, x);
right = Math.Max(right, x);
left = Math.Min(left, (int)x);
right = Math.Max(right, (int)x);
}
}

Expand Down