[FFmpeg-devel] [PATCH 6/6] lavc/ac3dsp: RISC-V B ac3_extract_exponents

2023-06-15 Thread Peiting Shen
From: Shen Peiting Use RVB instruction clz to calculate the number of leading zeros of MSB instead of av_log2. Benchmarks on Spike(cycles): ac3_extract_exponents_c: 8226 ac3_extract_exponents_rvb: 1167 Co-Authored by: Yang Xiaojun Co-Authored by: Huang Xing Co-Authored by: Zeng Fanchen Sign

[FFmpeg-devel] [PATCH 5/6] lavc/ac3dsp: RISC-V V ac3_compute_mantissa_size

2023-06-15 Thread Peiting Shen
From: Shen Peiting Use RVV instruction vlsege to operate on matrix columns. Benchmarks on Spike(cycles): ac3_compute_mantissa_size_c: 2338 ac3_compute_mantissa_size_rvv: 55 Co-Authored by: Yang Xiaojun Co-Authored by: Huang Xing Co-Authored by: Zeng Fanchen Signed-off-by: Shen Peiting ---

[FFmpeg-devel] [PATCH 4/6] lavc/ac3dsp: RISC-V V ac3_sum_square_butterfly_float

2023-06-15 Thread Peiting Shen
From: Shen Peiting Scalar calculating float sum_square optimized by using RVV instructions Benchmarks on Spike(cycles): len=128 ac3_sum_square_butterfly_float_c: 7986 ac3_sum_square_butterfly_float_rvv: 146 len=1280 ac3_sum_square_butterfly_float_c: 79410 ac3_sum_square_butterfly_float_rvv: 1154

[FFmpeg-devel] [PATCH 3/6] lavc/ac3dsp: RISC-V V ac3_sum_square_butterfly_int32

2023-06-15 Thread Peiting Shen
From: Shen Peiting Scalar calculating int32 sum_square optimized by using RVV instructions Benchmarks on Spike(cycles): len=128 ac3_sum_square_butterfly_int32_c: 8497 ac3_sum_square_butterfly_int32_rvv: 258 len=1280 ac3_sum_square_butterfly_int32_c: 84529 ac3_sum_square_butterfly_int32_rvv: 2274

[FFmpeg-devel] [PATCH 2/6] lavc/ac3dsp: RISC-V V float_to_fixed24

2023-06-15 Thread Peiting Shen
From: Shen Peiting Vector instructions replaces scalar options of float convert to fixed Benchmarks on Spike(cycles): len=16 float_to_fixed24_c: 315 float_to_fixed24_rvv: 27 len=160 float_to_fixed24_c: 2871 float_to_fixed24_rvv: 67 Co-Authored by: Yang Xiaojun Co-Authored by: Huang Xing Co-Au

[FFmpeg-devel] [PATCH 1/6] lavc/ac3dsp: RISC-V V ac3_exponent_min

2023-06-15 Thread Peiting Shen
From: Shen Peiting Find scalar minium optimized by using RVV instructions Benchmarks on Spike(cycles): *exp=1280*4;num_reuse_blocks=5;nb_coefs=16 ac3_exponent_min_c: 1993 ac3_exponent_min_rvv: 258 *exp=1280*4;num_reuse_blocks=19;nb_coefs=255 ac3_exponent_min_c: 99010 ac3_exponent_min_rvv: 3843

[FFmpeg-devel] [PATCH 0/6] RISC-V initial ac3dsp

2023-06-15 Thread Peiting Shen
From: Shen Peiting We optimized the six interfaces of AC3 init by RVV, the optimized performance was tested on the RISC-V ISA simulator--Spike, and the results were attached to each commit. shenpeiting (6): lavc/ac3dsp: RISC-V V ac3_exponent_min lavc/ac3dsp: RISC-V V float_to_fixed24 lav