42   const v128_t vw89 = wasm_v64x2_load_splat(weights + 8);  in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_3x4()  local
144 …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_3x4()
145 …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_3x4()
146 …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_3x4()
183 …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_3x4()
184 …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_3x4()
185 …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_3x4()
244 …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_3x4()
245 …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_3x4()
246 …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_3x4()
276 …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_3x4()
277 …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_3x4()
278 …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_3x4()