40   const v128_t vw89AB = wasm_v128_load(weights + 8);  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()  local
124 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
125 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
126 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
127 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
205 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
206 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi3x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
207 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi4x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
208 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi5x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
234 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
235 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
236 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
237 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
276 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
277 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
278 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
279 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
341 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
342 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
343 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
344 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
422 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
423 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi3x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
424 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi4x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
425 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi5x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
451 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
452 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
453 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
454 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
493 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
494 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
495 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
496 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
552 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
553 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
554 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
555 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x4567, wasm_v32x4_shuffle(vw89AB, vw89AB, 0, 0, 0,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
625 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi2x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
626 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi3x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
627 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi4x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
628 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi5x2345, wasm_v32x4_shuffle(vw89AB, vw89AB, 3, 3, 3,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
654 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
655 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
656 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
657 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x5678, wasm_v32x4_shuffle(vw89AB, vw89AB, 1, 1, 1,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
688 …vo0p0 = wasm_f32x4_add(vo0p0, wasm_f32x4_mul(vi1x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
689 …vo1p0 = wasm_f32x4_add(vo1p0, wasm_f32x4_mul(vi2x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
690 …vo2p0 = wasm_f32x4_add(vo2p0, wasm_f32x4_mul(vi3x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()
691 …vo3p0 = wasm_f32x4_add(vo3p0, wasm_f32x4_mul(vi4x6789, wasm_v32x4_shuffle(vw89AB, vw89AB, 2, 2, 2,…  in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_4x4()