Patryk Darmofalski b3b3b1bfe4 avutil/aarch64/float_dsp: add NEON scalarproduct_double ...

Four accumulators, as in scalarproduct_float. Eight were not faster at len 256.

checkasm --bench on Apple M4:
scalarproduct_double_c:        80.3
scalarproduct_double_neon:     21.8 ( 3.68x)

Signed-off-by: Patryk Darmofalski <darpat039@gmail.com>

2026-11-03 18:27:30 +00:00

..

aes_init.c

avutil/aes: add aarch64 NEON implementation

2026-09-16 06:29:59 +00:00

aes.S

avutil/aes: add aarch64 NEON implementation

2026-09-16 06:29:59 +00:00

asm.S

avutil/cpu: add aarch64 CPU feature flag for ARM AES

2026-09-16 06:29:59 +00:00

base64.h

avutil/base64: add aarch64 NEON implementation for base64 encode

2026-09-18 14:20:59 +00:00

base64.S

avutil/base64: add aarch64 NEON implementation for base64 encode

2026-09-18 14:20:59 +00:00

cpu_sme.S

configure: add detection of assembler support for SME

2025-12-09 21:38:38 +00:00

cpu_sve.S

aarch64: Print the SVE vector length in libavutil/tests/cpu.c

2024-09-27 00:06:30 +03:00

cpu.c

avutil/cpu: add aarch64 CPU feature flag for ARM AES

2026-09-16 06:29:59 +00:00

cpu.h

avutil/cpu: add aarch64 CPU feature flag for ARM AES

2026-09-16 06:29:59 +00:00

crc.h

avutil/crc: add aarch64 hybrid crc32 NEON PMULL+EOR SIMD implementation

2026-07-02 09:03:25 +00:00

crc.S

avutil/crc: add aarch64 hybrid crc32 NEON PMULL+EOR SIMD implementation

2026-07-02 09:03:25 +00:00

float_dsp_init.c

avutil/aarch64/float_dsp: add NEON scalarproduct_double

2026-11-03 18:27:30 +00:00

float_dsp_neon.S

avutil/aarch64/float_dsp: add NEON scalarproduct_double

2026-11-03 18:27:30 +00:00

intreadwrite.h

avutil/aarch64: add AV_COPY128 and AV_ZERO128 macros

2024-08-26 10:26:59 +02:00

Makefile

avutil/base64: add aarch64 NEON implementation for base64 encode

2026-09-18 14:20:59 +00:00

neontest.h

…

pixelutils_neon.S

avutil/aarch64: add pixelutils 32x32 SAD NEON implementation

2026-04-19 19:27:55 +00:00

pixelutils.h

avutil/aarch64: add pixelutils 32x32 SAD NEON implementation

2026-04-19 19:27:55 +00:00

timer.h

arm, aarch64: Don't use the ccnt / pmccntr_el0 registers for READ_TIME

2026-09-30 17:16:00 +00:00

tx_float_init.c

lavu/tx: contiguous fast path for the AArch64 inverse MDCT pre-rotation

2026-10-04 17:38:35 +09:00

tx_float_neon.S

lavu/tx: contiguous fast path for the AArch64 inverse MDCT pre-rotation

2026-10-04 17:38:35 +09:00