42   const v128_t vw89 = wasm_v64x2_load_splat(weights + 8);  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()  local
169 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
170 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi4x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
171 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi6x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
172 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi8x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
218 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
219 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi4x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
220 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi6x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
221 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi8x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
294 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
295 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi4x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
296 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi6x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
297 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi8x8ACE, wasm_v32x4_shuffle(vw89, vw89, 0, 0, 0, 0))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
334 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
335 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi4x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
336 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi6x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()
337 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi8x9BDF, wasm_v32x4_shuffle(vw89, vw89, 1, 1, 1, 1))…  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_4x4()