mirror of
https://git.ffmpeg.org/ffmpeg.git
synced 2025-12-23 23:40:15 +01:00
Accept up to 13 ULP difference. This fixes running "checkasm --test=ac3dsp 3044836819" on ARM. Depending on how the SIMD implementations aggregate numbers, larger/smaller values might not end up accumulated in exactly the same way; the current NEON implementation for ARM aggregates into vectors of 2 elements. If it would aggregate into vectors of 4 elements instead, like the AArch64 version does, this particular case would end up with a smaller difference.
5.8 KiB
5.8 KiB