DROOdotFOO ff22978529 avfilter/aarch64: NEON threshold filter ...

depth==8 routes to threshold8_neon (16 bytes/chunk); depth > 8 to
threshold16_neon (8 shorts/chunk, covering 9/10/12/14/16). C is not
auto-vectorized: four independent input streams defeat clang's
vectorizer. Per chunk the NEON kernel is cmhs+bsl.

checkasm --bench, Apple M1 (clang), speedup vs C:

threshold8    15.0x
threshold10    8.1x
threshold12    8.2x
threshold16    8.2x

Signed-off-by: DROOdotFOO <drew@axol.io>

2026-11-04 20:14:23 +00:00

..

Makefile

avfilter/aarch64: NEON threshold filter

2026-11-04 20:14:23 +00:00

vf_bwdif_init_aarch64.c

avfilter/aarch64/vf_bwdif_init: Fix stride bound check

2026-08-19 20:06:22 +00:00

vf_bwdif_neon.S

…

vf_colordetect_init.c

avfilter/vf_colordetect: Rename header to vf_colordetectdsp.h

2025-09-16 18:22:24 +02:00

vf_colordetect_neon.S

avfilter/vf_colordetect: add support for full range alpha offset

2026-09-01 13:15:06 +02:00

vf_nlmeans_init.c

lavfi/nlmeans: add aarch64 neon for compute_weights_line

2026-01-09 16:10:10 +00:00

vf_nlmeans_neon.S

lavfi/nlmeans: add aarch64 neon for compute_weights_line

2026-01-09 16:10:10 +00:00

vf_threshold_init.c

avfilter/aarch64: NEON threshold filter

2026-11-04 20:14:23 +00:00

vf_threshold_neon.S

avfilter/aarch64: NEON threshold filter

2026-11-04 20:14:23 +00:00