diff options
| author | Rémi Denis-Courmont <remi@remlab.net> | 2022-09-26 17:52:33 +0300 |
|---|---|---|
| committer | Lynne <dev@lynne.ee> | 2022-09-27 13:19:52 +0200 |
| commit | d120ab5b915fb422247f9e4dd5b823d1b47d6adf (patch) | |
| tree | a7783c549a246b745a9d35a8fac7d6887a5abbb4 /libavutil/riscv | |
| parent | c3db27ba9552fc3cb764bf9eb717840d981c5779 (diff) | |
| download | ffmpeg-d120ab5b915fb422247f9e4dd5b823d1b47d6adf.tar.gz | |
lavu/floatdsp: RISC-V V vector_dmac_scalar
Diffstat (limited to 'libavutil/riscv')
| -rw-r--r-- | libavutil/riscv/float_dsp_init.c | 3 | ||||
| -rw-r--r-- | libavutil/riscv/float_dsp_rvv.S | 18 |
2 files changed, 21 insertions, 0 deletions
diff --git a/libavutil/riscv/float_dsp_init.c b/libavutil/riscv/float_dsp_init.c index 9e19413d5d..a559bbb32b 100644 --- a/libavutil/riscv/float_dsp_init.c +++ b/libavutil/riscv/float_dsp_init.c @@ -34,6 +34,8 @@ void ff_vector_fmul_scalar_rvv(float *dst, const float *src, float mul, void ff_vector_dmul_rvv(double *dst, const double *src0, const double *src1, int len); +void ff_vector_dmac_scalar_rvv(double *dst, const double *src, double mul, + int len); void ff_vector_dmul_scalar_rvv(double *dst, const double *src, double mul, int len); @@ -50,6 +52,7 @@ av_cold void ff_float_dsp_init_riscv(AVFloatDSPContext *fdsp) if (flags & AV_CPU_FLAG_RVV_F64) { fdsp->vector_dmul = ff_vector_dmul_rvv; + fdsp->vector_dmac_scalar = ff_vector_dmac_scalar_rvv; fdsp->vector_dmul_scalar = ff_vector_dmul_scalar_rvv; } #endif diff --git a/libavutil/riscv/float_dsp_rvv.S b/libavutil/riscv/float_dsp_rvv.S index 4c325db9fd..048ec0bc40 100644 --- a/libavutil/riscv/float_dsp_rvv.S +++ b/libavutil/riscv/float_dsp_rvv.S @@ -91,6 +91,24 @@ func ff_vector_dmul_rvv, zve64d ret endfunc +// (a0) += (a1) * fa0 [0..a2-1] +func ff_vector_dmac_scalar_rvv, zve64d +NOHWD fmv.d.x fa0, a2 +NOHWD mv a2, a3 +1: + vsetvli t0, a2, e64, m1, ta, ma + vle64.v v24, (a1) + sub a2, a2, t0 + vle64.v v16, (a0) + sh3add a1, t0, a1 + vfmacc.vf v16, fa0, v24 + vse64.v v16, (a0) + sh3add a0, t0, a0 + bnez a2, 1b + + ret +endfunc + // (a0) = (a1) * fa0 [0..a2-1] func ff_vector_dmul_scalar_rvv, zve64d NOHWD fmv.d.x fa0, a2 |