diff --git a/paddle/fluid/operators/p_norm_op.cu b/paddle/fluid/operators/p_norm_op.cu
index 11a4928fe1c39cf75686cebd32451fdaec4d425b..1db6f6e51746282f89e9c160a0d370b72eecd605 100644
--- a/paddle/fluid/operators/p_norm_op.cu
+++ b/paddle/fluid/operators/p_norm_op.cu
@@ -22,6 +22,7 @@ namespace cub = hipcub;
 #endif
 #include "paddle/fluid/operators/amp/fp16_type_traits.h"
 #include "paddle/fluid/operators/elementwise/elementwise_op_impl.cu.h"
+#include "paddle/fluid/operators/fc_op.h"
 #include "paddle/fluid/operators/p_norm_op.h"
 #include "paddle/fluid/operators/reduce_ops/reduce_op.cu.h"
 #include "paddle/fluid/operators/reduce_ops/reduce_op.h"
@@ -106,9 +107,10 @@ class PnormCUDAKernel : public framework::OpKernel<T> {
     auto xdim = in_x->dims();
     auto ndim = out_norm->dims();
     float porder = ctx.Attr<float>("porder");
+    bool asvector = ctx.Attr<bool>("asvector");
     int axis = ctx.Attr<int>("axis");
-    if (axis < 0) axis = xdim.size() + axis;
     std::vector<int> reduce_axis = {axis};
+    reduce_axis = GetReduceDim(reduce_axis, xdim.size(), asvector);
 
     auto stream = ctx.cuda_device_context().stream();
 
@@ -195,7 +197,7 @@ class PnormGradCUDAKernel : public framework::OpKernel<T> {
     auto xdim = in_x->dims();
     float porder = ctx.Attr<float>("porder");
     int axis = ctx.Attr<int>("axis");
-    bool reduce_all = ((axis < 0) || (in_norm->numel() == 1));
+    bool reduce_all = (in_norm->numel() == 1);
     if (axis < 0) axis = xdim.size() + axis;
     const std::vector<int> dims = {axis};