/external/XNNPACK/src/f16-gemm/gen/ |
D | 4x8-minmax-neonfp16arith-ld64.c | 122 const float16x8_t va3c2 = vdupq_lane_f16(va3, 2); in xnn_f16_gemm_minmax_ukernel_4x8__neonfp16arith_ld64() local
|
/external/XNNPACK/src/qs8-gemm/gen/ |
D | 4x8c2-minmax-rndnu-neon-mull-ld1r.c | 145 const int8x8_t va3c2 = vreinterpret_s8_s16(va32); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld1r() local 262 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld1r() local
|
D | 4x8c2-minmax-rndnu-neon-mull-ld4r.c | 133 const int8x8_t va3c2 = vreinterpret_s8_s16(va3.val[2]); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld4r() local 250 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld4r() local
|
D | 4x8c2-minmax-rndnu-neon-mull-dup.c | 133 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_dup() local 250 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_dup() local
|
D | 4x8c2-minmax-rndnu-neon-mull-ld2r.c | 137 const int8x8_t va3c2 = vreinterpret_s8_s16(va31.val[0]); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld2r() local 254 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld2r() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 4x8c2-minmax-rndnu-neon-mull-ld4r.c | 150 const int8x8_t va3c2 = vreinterpret_s8_s16(va3.val[2]); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld4r() local 267 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld4r() local
|
D | 4x8c2-minmax-rndnu-neon-mull-ld2r.c | 154 const int8x8_t va3c2 = vreinterpret_s8_s16(va31.val[0]); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld2r() local 271 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld2r() local
|
D | 4x8c2-minmax-rndnu-neon-mull-ld1r.c | 162 const int8x8_t va3c2 = vreinterpret_s8_s16(va32); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld1r() local 279 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_ld1r() local
|
D | 4x8c2-minmax-rndnu-neon-mull-dup.c | 150 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_dup() local 267 const int8x8_t va3c2 = vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_s8(va3), 2)); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2__neon_mull_dup() local
|
/external/XNNPACK/src/f16-igemm/gen/ |
D | 4x8-minmax-neonfp16arith-ld64.c | 142 const float16x8_t va3c2 = vdupq_lane_f16(va3, 2); in xnn_f16_igemm_minmax_ukernel_4x8__neonfp16arith_ld64() local
|
/external/XNNPACK/src/f32-gemm/gen/ |
D | 4x8-relu-wasmrelaxedsimd-fma-splat.c | 116 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemm_relu_ukernel_4x8__wasmrelaxedsimd_fma_splat() local
|
D | 4x8-wasmsimd-splat.c | 116 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemm_ukernel_4x8__wasmsimd_splat() local
|
D | 4x8-wasmrelaxedsimd-fma-splat.c | 116 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemm_ukernel_4x8__wasmrelaxedsimd_fma_splat() local
|
D | 4x8-relu-wasmsimd-splat.c | 116 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemm_relu_ukernel_4x8__wasmsimd_splat() local
|
D | 4x8-minmax-wasmsimd-arm-splat.c | 118 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemm_minmax_ukernel_4x8__wasmsimd_arm_splat() local
|
/external/XNNPACK/src/f16-gemm/gen-inc/ |
D | 4x8inc-minmax-neonfp16arith-ld64.c | 124 const float16x8_t va3c2 = vdupq_lane_f16(va3, 2); in xnn_f16_gemminc_minmax_ukernel_4x8__neonfp16arith_ld64() local
|
/external/XNNPACK/src/f32-igemm/gen/ |
D | 4x8-relu-wasmrelaxedsimd-fma-splat.c | 138 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_igemm_relu_ukernel_4x8__wasmrelaxedsimd_fma_splat() local
|
D | 4x8-minmax-wasmsimd-arm-splat.c | 140 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_igemm_minmax_ukernel_4x8__wasmsimd_arm_splat() local
|
D | 4x8-wasmrelaxedsimd-fma-splat.c | 138 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_igemm_ukernel_4x8__wasmrelaxedsimd_fma_splat() local
|
D | 4x8-relu-wasmsimd-splat.c | 138 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_igemm_relu_ukernel_4x8__wasmsimd_splat() local
|
D | 4x8-minmax-wasmrelaxedsimd-fma-splat.c | 140 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_igemm_minmax_ukernel_4x8__wasmrelaxedsimd_fma_splat() local
|
D | 4x8-wasmsimd-splat.c | 138 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_igemm_ukernel_4x8__wasmsimd_splat() local
|
/external/XNNPACK/src/f32-gemm/gen-inc/ |
D | 4x8inc-minmax-wasmrelaxedsimd-splat.c | 120 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemminc_minmax_ukernel_4x8__wasmrelaxedsimd_splat() local
|
D | 4x8inc-minmax-relaxedwasmsimd-splat.c | 120 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemminc_minmax_ukernel_4x8__relaxedwasmsimd_splat() local
|
D | 4x8inc-minmax-wasmsimd-x86-splat.c | 120 const v128_t va3c2 = wasm_v32x4_shuffle(va3, va3, 2, 2, 2, 2); in xnn_f32_gemminc_minmax_ukernel_4x8__wasmsimd_x86_splat() local
|