339         vst1_f32(o1c0, vget_low_f32(vo1c0123.val[0])); o1c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
340         vst1_f32(o1c1, vget_high_f32(vo1c0123.val[0])); o1c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
341         vst1_f32(o1c2, vget_low_f32(vo1c0123.val[1])); o1c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
342         vst1_f32(o1c3, vget_high_f32(vo1c0123.val[1])); o1c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
344         vst1_f32(o0c0, vget_low_f32(vo0c0123.val[0])); o0c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
345         vst1_f32(o0c1, vget_high_f32(vo0c0123.val[0])); o0c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
346         vst1_f32(o0c2, vget_low_f32(vo0c0123.val[1])); o0c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
347         vst1_f32(o0c3, vget_high_f32(vo0c0123.val[1])); o0c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
597           vst1_f32(o1c0, vget_low_f32(vo1c0123.val[0])); o1c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
598           vst1_f32(o1c1, vget_high_f32(vo1c0123.val[0])); o1c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
599           vst1_f32(o1c2, vget_low_f32(vo1c0123.val[1])); o1c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
600           vst1_f32(o1c3, vget_high_f32(vo1c0123.val[1])); o1c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
602           vst1_f32(o0c0, vget_low_f32(vo0c0123.val[0])); o0c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
603           vst1_f32(o0c1, vget_high_f32(vo0c0123.val[0])); o0c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
604           vst1_f32(o0c2, vget_low_f32(vo0c0123.val[1])); o0c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()
605           vst1_f32(o0c3, vget_high_f32(vo0c0123.val[1])); o0c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2()