/external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
D | 3x3s2p1-minmax-neon-2x4.c | 106 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_2x4() local 173 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_2x4() local
|
D | 3x3s2p1-minmax-neonfma-2x4.c | 106 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4() local 173 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4() local
|
D | 3x3s2p1-minmax-neonfma-2x4-acc2.c | 106 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4_acc2() local 175 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4_acc2() local
|
D | 3x3s2p1-minmax-neonfma-3x4.c | 124 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_3x4() local 214 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_3x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-2x4.c | 130 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_2x4() local 201 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_2x4() local
|
D | 3x3s2p1-minmax-neon-3x4.c | 124 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_3x4() local 214 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_3x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-2x4.c | 130 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_2x4() local 201 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_2x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-loadsplat-2x4.c | 140 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_2x4() local 211 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_2x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-2x4-acc2.c | 130 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_2x4_acc2() local 203 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_2x4_acc2() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-2x4-acc2.c | 130 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_2x4_acc2() local 203 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_2x4_acc2() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-loadsplat-2x4.c | 140 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_2x4() local 211 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_2x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-loadsplat-2x4-acc2.c | 140 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_2x4_acc2() local 213 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_2x4_acc2() local
|
D | 3x3s2p1-minmax-sse-2x4.c | 138 const __m128 vi4x7BDF = _mm_move_ss(vi4xF9BD, vi4x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_2x4() local 214 const __m128 vi4x7BDF = _mm_move_ss(vi4xF9BD, vi4x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_2x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-loadsplat-2x4-acc2.c | 140 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_2x4_acc2() local 213 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_2x4_acc2() local
|
D | 3x3s2p1-minmax-sse-2x4-acc2.c | 138 const __m128 vi4x7BDF = _mm_move_ss(vi4xF9BD, vi4x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_2x4_acc2() local 216 const __m128 vi4x7BDF = _mm_move_ss(vi4xF9BD, vi4x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_2x4_acc2() local
|
D | 3x3s2p1-minmax-neonfma-4x4.c | 142 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_4x4() local 255 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_4x4() local
|
D | 3x3s2p1-minmax-neon-4x4.c | 142 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_4x4() local 255 const float32x4_t vi4x7BDF = vextq_f32(vi4x1357, vi4x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_4x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-3x4.c | 156 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_3x4() local 252 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_3x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-3x4.c | 156 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_3x4() local 252 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_3x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-loadsplat-3x4.c | 166 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_3x4() local 262 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_3x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-loadsplat-3x4.c | 166 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_3x4() local 262 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_3x4() local
|
D | 3x3s2p1-minmax-sse-3x4.c | 168 const __m128 vi4x7BDF = _mm_move_ss(vi4xF9BD, vi4x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_3x4() local 271 const __m128 vi4x7BDF = _mm_move_ss(vi4xF9BD, vi4x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_3x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-4x4.c | 182 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4() local 303 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-4x4.c | 182 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_4x4() local 303 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_4x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-loadsplat-4x4.c | 192 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_4x4() local 313 const v128_t vi4x7BDF = wasm_v32x4_shuffle(vi4x1357, vi4x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_4x4() local
|