diff options
| author | Hubert Mazur <hum@semihalf.com> | 2022-09-20 13:01:57 +0200 |
|---|---|---|
| committer | Martin Storsjö <martin@martin.st> | 2022-09-21 12:57:56 +0300 |
| commit | e9a6170213d9fc48db4f4135b9fd848f461a0bd2 (patch) | |
| tree | f7c6116e3745ed8e8a2bad6d1c0eca9370322d71 /libavcodec/aarch64/me_cmp_neon.S | |
| parent | 0ee535b1db392a340a86ce0581dff94624c2da06 (diff) | |
| download | ffmpeg-e9a6170213d9fc48db4f4135b9fd848f461a0bd2.tar.gz | |
lavc/aarch64: Add neon implementation for vsad8_intra
Provide optimized implementation for vsad8_intra function.
Performance comparison tests are shown below.
- vsad_5_c: 94.7
- vsad_5_neon: 20.7
Benchmarks and tests run with checkasm tool on AWS Graviton 3.
Signed-off-by: Hubert Mazur <hum@semihalf.com>
Signed-off-by: Martin Storsjö <martin@martin.st>
Diffstat (limited to 'libavcodec/aarch64/me_cmp_neon.S')
| -rw-r--r-- | libavcodec/aarch64/me_cmp_neon.S | 42 |
1 files changed, 42 insertions, 0 deletions
diff --git a/libavcodec/aarch64/me_cmp_neon.S b/libavcodec/aarch64/me_cmp_neon.S index 1e08d04869..1587f2ea43 100644 --- a/libavcodec/aarch64/me_cmp_neon.S +++ b/libavcodec/aarch64/me_cmp_neon.S @@ -1047,3 +1047,45 @@ function pix_median_abs16_neon, export=1 ret endfunc + +function vsad_intra8_neon, export=1 + // x0 unused + // x1 uint8_t *pix1 + // x2 uint8_t *dummy + // x3 ptrdiff_t stride + // w4 int h + + ld1 {v0.8b}, [x1], x3 + sub w4, w4, #1 // we need to make h-1 iterations + cmp w4, #3 + movi v16.8h, #0 + b.lt 2f + +1: + // v = abs( pix1[0] - pix1[0 + stride] ) + // score = sum(v) + ld1 {v1.8b}, [x1], x3 + sub w4, w4, #3 + ld1 {v2.8b}, [x1], x3 + uabal v16.8h, v0.8b, v1.8b + ld1 {v3.8b}, [x1], x3 + uabal v16.8h, v1.8b, v2.8b + cmp w4, #3 + mov v0.8b, v3.8b + uabal v16.8h, v2.8b, v3.8b + b.ge 1b + cbz w4, 3f + +2: + ld1 {v1.8b}, [x1], x3 + subs w4, w4, #1 + uabal v16.8h, v0.8b, v1.8b + mov v0.8b, v1.8b + cbnz w4, 2b + +3: + uaddlv s17, v16.8h + fmov w0, s17 + + ret +endfunc |