lodepng-turbo is a fast PNG image codec that uses SIMD instructions (MMX, SSE2, AVX2, NEON) to accelerate baseline PNG decompression on x86, x86-64, ARM systems.
At the stage of writing this issue, lodepng-turbo 's unfilter process is not working properly.
This is a problem of the behavior of the asm function of SIMD implemented in parng.
parng is not a C++ program, it is unknown how to correctly call each function implemented by asm.
Alternatively, there is a possibility of being an unfinished implementation.
There are several options to fix this problem.
Modify the calling method and make it operate properly
Switch to another implementation such as libpng
New implementation of unfilter with asm (it will be a C ++ program using asm syntax)
Implement unfilter with Halide
use depng (Since buffer handling is incomplete, improvement is necessary)