What is the reason of using 256 for prefetchnta?
https://github.com/skywind3000/FastMemcpy/blob/8fea5f666be174c6548d0ae4010e81b0a742c853/FastMemcpy.h#L644
Hi, it bugs me as 128 seems to be a reasonable choice. Is that derived from experiments? Or something related to the mechanism of prefetching itself?
https://github.com/skywind3000/FastMemcpy/blob/8fea5f666be174c6548d0ae4010e81b0a742c853/FastMemcpy.h#L644
Hi, it bugs me as 128 seems to be a reasonable choice. Is that derived from experiments? Or something related to the mechanism of prefetching itself?
It bugs me too. But I guess it's just an experimental value about how far to look ahead.
https://github.com/skywind3000/FastMemcpy/blob/8fea5f666be174c6548d0ae4010e81b0a742c853/FastMemcpy.h#L644
Hi, it bugs me as 128 seems to be a reasonable choice. Is that derived from experiments? Or something related to the mechanism of prefetching itself?
It bugs me too. But I guess it's just an experimental value about how far to look ahead.
I don't quite understand what the experimental value is in this case. Could you give an example?