/external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
D | 3x3s2p1-minmax-neonfma-1x4.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4() local 127 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4() local
|
D | 3x3s2p1-minmax-neon-1x4.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4() local 127 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4() local
|
D | 3x3s2p1-minmax-neonfma-1x4-acc2.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4_acc2() local 128 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4_acc2() local
|
D | 3x3s2p1-minmax-neon-1x4-acc2.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4_acc2() local 128 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4_acc2() local
|
D | 3x3s2p1-minmax-neon-1x4-acc3.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4_acc3() local 129 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4_acc3() local
|
D | 3x3s2p1-minmax-neonfma-1x4-acc3.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4_acc3() local 129 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4_acc3() local
|
D | 3x3s2p1-minmax-neonfma-1x4-acc4.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4_acc4() local 130 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_1x4_acc4() local
|
D | 3x3s2p1-minmax-neon-1x4-acc4.c | 79 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4_acc4() local 130 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4_acc4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-1x4.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4() local 145 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-1x4.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4() local 145 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-1x4-acc3.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc3() local 147 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc3() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-1x4-acc2.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc2() local 146 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc2() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-1x4-acc2.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc2() local 146 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc2() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-1x4-acc3.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc3() local 147 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc3() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-splat-1x4-acc4.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc4() local 148 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-splat-1x4-acc4.c | 95 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc4() local 148 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-loadsplat-1x4.c | 105 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_1x4() local 155 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_1x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-loadsplat-1x4-acc2.c | 105 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_1x4_acc2() local 156 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_1x4_acc2() local
|
D | 3x3s2p1-minmax-neon-2x4.c | 98 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_2x4() local 169 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_2x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-loadsplat-1x4-acc2.c | 105 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_1x4_acc2() local 156 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_1x4_acc2() local
|
D | 3x3s2p1-minmax-sse-1x4.c | 103 const __m128 vi0x7BDF = _mm_move_ss(vi0xF9BD, vi0x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_1x4() local 152 const __m128 vi0x7BDF = _mm_move_ss(vi0xF9BD, vi0x7531); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_1x4() local
|
D | 3x3s2p1-minmax-neonfma-2x4.c | 98 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4() local 169 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4() local
|
D | 3x3s2p1-minmax-wasmsimd-x86-loadsplat-1x4.c | 105 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_1x4() local 155 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_loadsplat_1x4() local
|
D | 3x3s2p1-minmax-wasmsimd-arm-loadsplat-1x4-acc4.c | 105 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_1x4_acc4() local 158 const v128_t vi0x7BDF = wasm_v32x4_shuffle(vi0x1357, vi0x9BDF, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_loadsplat_1x4_acc4() local
|
D | 3x3s2p1-minmax-neonfma-2x4-acc2.c | 98 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x8ACE9BDF.val[1], 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4_acc2() local 171 const float32x4_t vi0x7BDF = vextq_f32(vi0x1357, vi0x9BDF, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4_acc2() local
|