mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2025-02-04 06:08:26 +02:00
4d66e8c12e
The loop uses a 32-bit accumulator. The current code would only zero the lower 16 bits thereof.