341         vst1_f32(o1c0, vget_low_f32(vo1c01)); o1c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
342         vst1_f32(o1c1, vget_high_f32(vo1c01)); o1c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
343         vst1_f32(o1c2, vget_low_f32(vo1c23)); o1c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
344         vst1_f32(o1c3, vget_high_f32(vo1c23)); o1c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
346         vst1_f32(o0c0, vget_low_f32(vo0c01)); o0c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
347         vst1_f32(o0c1, vget_high_f32(vo0c01)); o0c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
348         vst1_f32(o0c2, vget_low_f32(vo0c23)); o0c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
349         vst1_f32(o0c3, vget_high_f32(vo0c23)); o0c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
601           vst1_f32(o1c0, vget_low_f32(vo1c01)); o1c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
602           vst1_f32(o1c1, vget_high_f32(vo1c01)); o1c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
603           vst1_f32(o1c2, vget_low_f32(vo1c23)); o1c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
604           vst1_f32(o1c3, vget_high_f32(vo1c23)); o1c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
606           vst1_f32(o0c0, vget_low_f32(vo0c01)); o0c0 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
607           vst1_f32(o0c1, vget_high_f32(vo0c01)); o0c1 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
608           vst1_f32(o0c2, vget_low_f32(vo0c23)); o0c2 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()
609           vst1_f32(o0c3, vget_high_f32(vo0c23)); o0c3 += 2;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2()