AVX and AVX512F versions of PReLU micro-kernel

Up to 2X faster PReLU operator on AVX512F-enabled CPU

PiperOrigin-RevId: 300268958
diff --git a/test/f32-prelu.yaml b/test/f32-prelu.yaml
index fbdeb42..6faf6bf 100644
--- a/test/f32-prelu.yaml
+++ b/test/f32-prelu.yaml
@@ -8,6 +8,10 @@
 - name: xnn_f32_prelu_ukernel__sse2_2x8
 - name: xnn_f32_prelu_ukernel__sse41_2x4
 - name: xnn_f32_prelu_ukernel__sse41_2x8
+- name: xnn_f32_prelu_ukernel__avx_2x8
+- name: xnn_f32_prelu_ukernel__avx_2x16
+- name: xnn_f32_prelu_ukernel__avx512f_2x16
+- name: xnn_f32_prelu_ukernel__avx512f_2x32
 - name: xnn_f32_prelu_ukernel__psimd_2x4
 - name: xnn_f32_prelu_ukernel__psimd_2x8
 - name: xnn_f32_prelu_ukernel__scalar_2x1