tree 9ef63cc40499d9da9e17d4f15db4a96bbf7084b9
parent b64ea1171bfa7424d2e3b2ffba50c58fa15769a7
author Martijn van Beurden <mvanb1@gmail.com> 1652033676 +0200
committer GitHub <noreply@github.com> 1652033676 +0200
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJieAiMCRBK7hj4Ov3rIwAAHaAIAEiwB1JuK2oY7XSNBTZLRflA
 7IxMKNumhMr1wA0YcGlHupfztiwDkjyRSDi09g62Jdl+pvcner5A1IO5VDcKBsPH
 bZH20rr7lP0Vo3WedXkSInJ3fydDAr3eGEuXMOmxO/uhxn47LZcCfN6IboqjJh3J
 wQKxQrW8kKKDF41AUEX100/zCAR8V7VSFQgR3YUn6bxWh9hTs/DbBPzNPL0bHnjS
 ph2osuoImCBzNNydrZSiJ/wnWehxgQ+7KROXSe+2b+sYovYYBTYPSstgZijMBWaJ
 6UadB7pK+DDBjG3u8qoB2vju7msBKVqcz+jBfIqbQ4yCHW23wKTvqdCKSMQ2C4E=
 =Cn0f
 -----END PGP SIGNATURE-----
 

Rewrite of fixed_compute_best_predictor functions

The code of fixed_compute_best_predictor was presumable once optimized,
but it appears a much more readable form is now equally fast or even faster,
see https://github.com/xiph/flac/pull/337 for details.