Use shared status object.

src/TensorFlowNET.Core/Eager/EagerOperation.cs

@@ -53,8 +53,7 @@ namespace Tensorflow.Eager
         {
             object value = null;
             byte isList = 0;
-            using var status = new Status();
-            var attrType = c_api.TFE_OpNameGetAttrType(tf.context, Name, attr_name, ref isList, status);
+            var attrType = c_api.TFE_OpNameGetAttrType(tf.context, Name, attr_name, ref isList, tf.status);
             switch (attrType)
             {
                 case TF_AttrType.TF_ATTR_BOOL:

src/TensorFlowNET.Core/Eager/EagerTensor.Creation.cs

@@ -22,13 +22,13 @@ namespace Tensorflow.Eager
 
         public EagerTensor(string value, string device_name) : base(value)
         {
-            EagerTensorHandle = c_api.TFE_NewTensorHandle(_handle, status);
+            EagerTensorHandle = c_api.TFE_NewTensorHandle(_handle, tf.status);
             Resolve();
         }
         
         public EagerTensor(NDArray value, string device_name) : base(value)
         {
-            EagerTensorHandle = c_api.TFE_NewTensorHandle(_handle, status);
+            EagerTensorHandle = c_api.TFE_NewTensorHandle(_handle, tf.status);
             Resolve();
         }
 
@@ -37,7 +37,7 @@ namespace Tensorflow.Eager
             _id = get_uid();
 
             if (_handle == IntPtr.Zero)
-                _handle = c_api.TFE_TensorHandleResolve(EagerTensorHandle, status);
+                _handle = c_api.TFE_TensorHandleResolve(EagerTensorHandle, tf.status);
 
             //print($"new Tensor {Id} {_handle.ToString("x16")}");
             //print($"new TensorHandle {Id} {EagerTensorHandle.ToString("x16")}");

src/TensorFlowNET.Core/Eager/EagerTensor.cs

@@ -8,26 +8,23 @@ namespace Tensorflow.Eager
 {
     public partial class EagerTensor : Tensor
     {
-        Status status = new Status();
         public IntPtr EagerTensorHandle;
-        public override string Device => c_api.StringPiece(c_api.TFE_TensorHandleDeviceName(EagerTensorHandle, status));
+        public override string Device => c_api.StringPiece(c_api.TFE_TensorHandleDeviceName(EagerTensorHandle, tf.status));
 
-        public override int rank => c_api.TFE_TensorHandleNumDims(EagerTensorHandle, status);
+        public override int rank => c_api.TFE_TensorHandleNumDims(EagerTensorHandle, tf.status);
 
         public static int GetRank(IntPtr handle)
         {
             var tfe_tensor_handle = c_api.TFE_EagerTensorHandle(handle);
-            using var status = new Status();
-            return c_api.TFE_TensorHandleNumDims(tfe_tensor_handle, status);
+            return c_api.TFE_TensorHandleNumDims(tfe_tensor_handle, tf.status);
         }
 
         public static int[] GetDims(IntPtr handle)
         {
             var tfe_tensor_handle = c_api.TFE_EagerTensorHandle(handle);
-            using var status = new Status();
-            var dims = new int[c_api.TFE_TensorHandleNumDims(tfe_tensor_handle, status)];
+            var dims = new int[c_api.TFE_TensorHandleNumDims(tfe_tensor_handle, tf.status)];
             for (int i = 0; i < dims.Length; i++)
-                dims[i] = c_api.TFE_TensorHandleDim(tfe_tensor_handle, i, status);
+                dims[i] = c_api.TFE_TensorHandleDim(tfe_tensor_handle, i, tf.status);
             return dims;
         }
 

src/TensorFlowNET.Core/Graphs/Graph.cs

@@ -512,7 +512,7 @@ namespace Tensorflow
 
         public TensorShape GetTensorShape(TF_Output output)
         {
-            var status = new Status();
+            var status = tf.status;
             var ndim = c_api.TF_GraphGetTensorNumDims(_handle, output, status);
             status.Check();
 

src/TensorFlowNET.Core/Operations/Operation.Input.cs

@@ -17,6 +17,7 @@
 using System;
 using System.Linq;
 using System.Runtime.InteropServices;
+using static Tensorflow.Binding;
 
 namespace Tensorflow
 {
@@ -30,11 +31,8 @@ namespace Tensorflow
         public int InputListLength(string name)
         {
             int num = 0;
-            using(var status = new Status())
-            {
-                num = c_api.TF_OperationInputListLength(_handle, name, status);
-                status.Check(true);
-            }
+            num = c_api.TF_OperationInputListLength(_handle, name, tf.status);
+            tf.status.Check(true);
             return num;
         }
         public int NumInputs => c_api.TF_OperationNumInputs(_handle);

src/TensorFlowNET.Core/Operations/Operation.Output.cs

@@ -28,12 +28,8 @@ namespace Tensorflow
 
         public int OutputListLength(string name)
         {
-            int num = 0;
-            using (var status = new Status())
-            {
-                num = c_api.TF_OperationOutputListLength(_handle, name, status);
-                status.Check(true);
-            }
+            int num = c_api.TF_OperationOutputListLength(_handle, name, tf.status);
+            tf.status.Check(true);
 
             return num;
         }

src/TensorFlowNET.Core/Operations/Operation.cs

@@ -20,6 +20,7 @@ using System.Collections.Generic;
 using System.IO;
 using System.Linq;
 using Tensorflow.Util;
+using static Tensorflow.Binding;
 
 namespace Tensorflow
 {
@@ -233,14 +234,13 @@ namespace Tensorflow
             AttrValue x = null;
 
             lock (Locks.ProcessWide)
-                using (var status = new Status())
-                using (var buf = new Buffer())
-                {
-                    c_api.TF_OperationGetAttrValueProto(_handle, name, buf, status);
-                    status.Check(true);
+            {
+                using var buf = new Buffer();
+                c_api.TF_OperationGetAttrValueProto(_handle, name, buf, tf.status);
+                tf.status.Check(true);
 
-                    x = AttrValue.Parser.ParseFrom(buf.MemoryBlock.Stream());
-                }
+                x = AttrValue.Parser.ParseFrom(buf.MemoryBlock.Stream());
+            }
 
             string oneof_value = x.ValueCase.ToString();
             if (string.IsNullOrEmpty(oneof_value))
@@ -295,11 +295,10 @@ namespace Tensorflow
             // after the c_api call next time _inputs is accessed 
             // the updated inputs are reloaded from the c_api
             lock (Locks.ProcessWide)
-                using (var status = new Status())
                 {
-                    c_api.UpdateEdge(_graph, output, input, status);
+                    c_api.UpdateEdge(_graph, output, input, tf.status);
                     //var updated_inputs = inputs;
-                    status.Check();
+                    tf.status.Check();
                 }
         }
 

src/TensorFlowNET.Core/Operations/array_ops.cs

@@ -43,30 +43,26 @@ namespace Tensorflow
                 allow_broadcast: false);
 
         public static Tensor zeros(TensorShape shape, TF_DataType dtype = TF_DataType.TF_FLOAT, string name = null)
-        {
-            dtype = dtype.as_base_dtype();
-            return tf_with(ops.name_scope(name, "zeros", shape), scope =>
+            => tf_with(ops.name_scope(name, "zeros", shape), scope =>
             {
+                dtype = dtype.as_base_dtype();
                 name = scope;
+                var shape_tensor = constant_op._tensor_shape_tensor_conversion_function(shape);
+                Tensor zeros = null;
                 switch (dtype)
                 {
-                    case TF_DataType.TF_BOOL:
-                        return _constant_if_small(false, shape, dtype, name);
                     case TF_DataType.TF_DOUBLE:
-                        return _constant_if_small(0.0D, shape, dtype, name);
+                        zeros = constant(0d);
+                        break;
                     case TF_DataType.TF_FLOAT:
-                        return _constant_if_small(0.0F, shape, dtype, name);
-                    case TF_DataType.TF_INT64:
-                        return _constant_if_small(0L, shape, dtype, name);
-                    case TF_DataType.TF_INT32:
-                        return _constant_if_small(0, shape, dtype, name);
-                    case TF_DataType.TF_INT8:
-                        return _constant_if_small<byte>(0, shape, dtype, name);
+                        zeros = constant(0f);
+                        break;
                     default:
-                        throw new TypeError("can't find type for zeros");
+                        zeros = constant(0);
+                        break;
                 }
+                return fill(shape_tensor, zeros, name: name);
             });
-        }
 
         public static Tensor boolean_mask<T1, T2>(T1 tensor, T2 mask, string name = "boolean_mask", int axis = 0)
         {

src/TensorFlowNET.Core/Sessions/BaseSession.cs

@@ -22,7 +22,7 @@ using System.Linq;
 using System.Numerics;
 using System.Text;
 using Google.Protobuf;
-using NumSharp.Backends;
+using static Tensorflow.Binding;
 using Tensorflow.Util;
 
 namespace Tensorflow
@@ -236,7 +236,7 @@ namespace Tensorflow
             // Ensure any changes to the graph are reflected in the runtime.
             _extend_graph();
 
-            var status = new Status();
+            var status = tf.status;
 
             var output_values = fetch_list.Select(x => IntPtr.Zero).ToArray();
 

src/TensorFlowNET.Core/Sessions/Session.cs

@@ -46,7 +46,7 @@ namespace Tensorflow
             lock (Locks.ProcessWide)
             {
                 var graph = c_api.TF_NewGraph();
-                var status = new Status();
+                using var status = new Status();
                 var opt = new SessionOptions();
 
                 var tags = new string[] {"serve"};
@@ -66,7 +66,6 @@ namespace Tensorflow
                     status.Check(true);
                 } catch (TensorflowException ex) when (ex.Message.Contains("Could not find SavedModel"))
                 {
-                    status = new Status();
                     sess = c_api.TF_LoadSessionFromSavedModel(opt,
                         IntPtr.Zero,
                         Path.GetFullPath(path),

src/TensorFlowNET.Core/Tensors/EagerTensorV2.cs

@@ -13,14 +13,12 @@ namespace Tensorflow
     public class EagerTensorV2 : DisposableObject, ITensor
     {
         IntPtr EagerTensorHandle;
-        public string Device => c_api.StringPiece(c_api.TFE_TensorHandleDeviceName(EagerTensorHandle, status));
-
-        static Status status = new Status();
+        public string Device => c_api.StringPiece(c_api.TFE_TensorHandleDeviceName(EagerTensorHandle, tf.status));
 
         public EagerTensorV2(IntPtr handle)
         {
             EagerTensorHandle = c_api.TFE_EagerTensorHandle(handle);
-            _handle = c_api.TFE_TensorHandleResolve(EagerTensorHandle, status);
+            _handle = c_api.TFE_TensorHandleResolve(EagerTensorHandle, tf.status);
         }
 
         public unsafe EagerTensorV2(NDArray nd, string device_name = "")
@@ -40,7 +38,7 @@ namespace Tensorflow
 
                     }, IntPtr.Zero);
 
-            EagerTensorHandle = c_api.TFE_NewTensorHandle(_handle, status);
+            EagerTensorHandle = c_api.TFE_NewTensorHandle(_handle, tf.status);
         }
 
         /*public unsafe EagerTensorV2(float[,] value)

src/TensorFlowNET.Core/Tensors/Tensor.Conversions.cs

@@ -21,6 +21,7 @@ using System.Globalization;
 using System.Runtime.CompilerServices;
 using System.Text;
 using NumSharp.Utilities;
+using static Tensorflow.Binding;
 
 namespace Tensorflow
 {
@@ -69,11 +70,8 @@ namespace Tensorflow
                         IntPtr stringStartAddress = IntPtr.Zero;
                         UIntPtr dstLen = UIntPtr.Zero;
 
-                        using (var status = new Status())
-                        {
-                            c_api.TF_StringDecode((byte*) this.buffer + 8, (UIntPtr) (this.bytesize), (byte**) &stringStartAddress, &dstLen, status);
-                            status.Check(true);
-                        }
+                        c_api.TF_StringDecode((byte*) this.buffer + 8, (UIntPtr) (this.bytesize), (byte**) &stringStartAddress, &dstLen, tf.status);
+                        tf.status.Check(true);
 
                         var dstLenInt = checked((int) dstLen);
                         var value = Encoding.UTF8.GetString((byte*) stringStartAddress, dstLenInt);

src/TensorFlowNET.Core/Tensors/Tensor.Creation.cs

@@ -451,7 +451,6 @@ namespace Tensorflow
         /// </summary>
         public unsafe Tensor(string str)
         {
-            var status = new Status();
             var buffer = Encoding.UTF8.GetBytes(str);
             var size = c_api.TF_StringEncodedSize((UIntPtr)buffer.Length);
             var handle = TF_AllocateTensor(TF_DataType.TF_STRING, IntPtr.Zero, 0, (UIntPtr)((ulong)size + 8));
@@ -460,9 +459,9 @@ namespace Tensorflow
             IntPtr tensor = c_api.TF_TensorData(handle);
             Marshal.WriteInt64(tensor, 0);
             fixed (byte* src = buffer)
-                c_api.TF_StringEncode(src, (UIntPtr)buffer.Length, (sbyte*)(tensor + sizeof(long)), size, status);
+                c_api.TF_StringEncode(src, (UIntPtr)buffer.Length, (sbyte*)(tensor + sizeof(long)), size, tf.status);
             _handle = handle;
-            status.Check(true);
+            tf.status.Check(true);
         }
 
         public unsafe Tensor(NDArray nd, TF_DataType? tensorDType = null)
@@ -483,10 +482,8 @@ namespace Tensorflow
                     IntPtr tensor = c_api.TF_TensorData(handle);
                     Marshal.WriteInt64(tensor, 0);
 
-                    var status = new Status();
-                    c_api.TF_StringEncode((byte*) nd.Unsafe.Address, bytesLength, (sbyte*) (tensor + sizeof(Int64)), size, status);
-
-                    status.Check(true);
+                    c_api.TF_StringEncode((byte*) nd.Unsafe.Address, bytesLength, (sbyte*) (tensor + sizeof(Int64)), size, tf.status);
+                    tf.status.Check(true);
                     _handle = handle;
                 } else
                 {
@@ -498,11 +495,10 @@ namespace Tensorflow
                     IntPtr tensor = c_api.TF_TensorData(handle);
                     Marshal.WriteInt64(tensor, 0);
 
-                    var status = new Status();
                     fixed (byte* src = buffer)
-                        c_api.TF_StringEncode(src, (UIntPtr) buffer.Length, (sbyte*) (tensor + sizeof(Int64)), size, status);
+                        c_api.TF_StringEncode(src, (UIntPtr) buffer.Length, (sbyte*) (tensor + sizeof(Int64)), size, tf.status);
 
-                    status.Check(true);
+                    tf.status.Check(true);
                     _handle = handle;
                 }
 
@@ -607,11 +603,10 @@ namespace Tensorflow
                 IntPtr tensor = c_api.TF_TensorData(handle);
                 Marshal.WriteInt64(tensor, 0);
 
-                var status = new Status();
                 fixed (byte* src = buffer)
-                    c_api.TF_StringEncode(src, (UIntPtr) buffer.Length, (sbyte*) (tensor + sizeof(Int64)), size, status);
+                    c_api.TF_StringEncode(src, (UIntPtr) buffer.Length, (sbyte*) (tensor + sizeof(long)), size, tf.status);
 
-                status.Check(true);
+                tf.status.Check(true);
                 return handle;
             }
 

src/TensorFlowNET.Core/Tensors/Tensor.Value.cs

@@ -3,7 +3,7 @@ using NumSharp.Backends;
 using NumSharp.Backends.Unmanaged;
 using NumSharp.Utilities;
 using System;
-using System.Collections.Generic;
+using static Tensorflow.Binding;
 using System.Runtime.InteropServices;
 using System.Text;
 
@@ -237,18 +237,15 @@ namespace Tensorflow
             var src = c_api.TF_TensorData(_handle);
             var srcLen = (IntPtr)(src.ToInt64() + (long)bytesize);
             src += (int)(size * 8);
-            using (var status = new Status())
+            for (int i = 0; i < buffer.Length; i++)
             {
-                for (int i = 0; i < buffer.Length; i++)
-                {
-                    IntPtr dst = IntPtr.Zero;
-                    UIntPtr dstLen = UIntPtr.Zero;
-                    var read = c_api.TF_StringDecode((byte*)src, (UIntPtr)(srcLen.ToInt64() - src.ToInt64()), (byte**)&dst, &dstLen, status);
-                    status.Check(true);
-                    buffer[i] = new byte[(int)dstLen];
-                    Marshal.Copy(dst, buffer[i], 0, buffer[i].Length);
-                    src += (int)read;
-                }
+                IntPtr dst = IntPtr.Zero;
+                UIntPtr dstLen = UIntPtr.Zero;
+                var read = c_api.TF_StringDecode((byte*)src, (UIntPtr)(srcLen.ToInt64() - src.ToInt64()), (byte**)&dst, &dstLen, tf.status);
+                tf.status.Check(true);
+                buffer[i] = new byte[(int)dstLen];
+                Marshal.Copy(dst, buffer[i], 0, buffer[i].Length);
+                src += (int)read;
             }
 
             var _str = new string[buffer.Length];

src/TensorFlowNET.Core/Tensors/Tensor.cs

@@ -22,7 +22,7 @@ using System.Globalization;
 using System.Linq;
 using System.Runtime.InteropServices;
 using System.Text;
-using System.Threading.Tasks;
+using static Tensorflow.Binding;
 using Tensorflow.Framework;
 
 namespace Tensorflow
@@ -109,11 +109,7 @@ namespace Tensorflow
 
                 if (_handle == IntPtr.Zero)
                 {
-                    using (var status = new Status())
-                    {
-                        c_api.TF_GraphGetTensorShape(op.graph, _as_tf_output(), dims, rank, status);
-                        status.Check();
-                    }
+                    c_api.TF_GraphGetTensorShape(op.graph, _as_tf_output(), dims, rank, tf.status);
                 }
                 else
                 {
@@ -126,15 +122,12 @@ namespace Tensorflow
 
             set
             {
-                using (var status = new Status())
-                {
-                    if (value == null)
-                        c_api.TF_GraphSetTensorShape(graph, _as_tf_output(), null, -1, status);
-                    else
-                        c_api.TF_GraphSetTensorShape(graph, _as_tf_output(), value.Select(Convert.ToInt64).ToArray(), value.Length, status);
+                if (value == null)
+                    c_api.TF_GraphSetTensorShape(graph, _as_tf_output(), null, -1, tf.status);
+                else
+                    c_api.TF_GraphSetTensorShape(graph, _as_tf_output(), value.Select(Convert.ToInt64).ToArray(), value.Length, tf.status);
 
-                    status.Check(true);
-                }
+                tf.status.Check(true);
             }
         }
 
@@ -178,13 +171,9 @@ namespace Tensorflow
             {
                 if (_handle == IntPtr.Zero)
                 {
-                    using (var status = new Status())
-                    {
-                        var output = _as_tf_output();
-                        int ndim = c_api.TF_GraphGetTensorNumDims(op.graph, output, status);
-                        status.Check();
-                        return ndim;
-                    }
+                    var output = _as_tf_output();
+                    int ndim = c_api.TF_GraphGetTensorNumDims(op.graph, output, tf.status);
+                    return ndim;
                 }
 
                 return c_api.TF_NumDims(_handle);

src/TensorFlowNET.Core/ops.cs

@@ -176,30 +176,29 @@ namespace Tensorflow
                         throw new NotImplementedException("_create_c_op");
                 }
 
-                using (var status = new Status())
+                var status = tf.status;
+                
+                // Add control inputs
+                foreach (var control_input in control_inputs)
+                    c_api.TF_AddControlInput(op_desc, control_input);
+
+                // Add attrs
+                foreach (var attr in node_def.Attr)
                 {
-                    // Add control inputs
-                    foreach (var control_input in control_inputs)
-                        c_api.TF_AddControlInput(op_desc, control_input);
-
-                    // Add attrs
-                    foreach (var attr in node_def.Attr)
-                    {
-                        var bytes = attr.Value.ToByteArray(); //TODO: we can use attr.Value.WriteTo with a memory stream.
-                        var protoHandle = Marshal.AllocHGlobal(bytes.Length);
-                        Marshal.Copy(bytes, 0, protoHandle, bytes.Length);
-                        uint len = (uint)bytes.Length;
-                        c_api.TF_SetAttrValueProto(op_desc, attr.Key, protoHandle, proto_len: len, status: status);
-                        status.Check(true);
-                        Marshal.FreeHGlobal(protoHandle);
-                    }
+                    var bytes = attr.Value.ToByteArray(); //TODO: we can use attr.Value.WriteTo with a memory stream.
+                    var protoHandle = Marshal.AllocHGlobal(bytes.Length);
+                    Marshal.Copy(bytes, 0, protoHandle, bytes.Length);
+                    uint len = (uint)bytes.Length;
+                    c_api.TF_SetAttrValueProto(op_desc, attr.Key, protoHandle, proto_len: len, status: status);
+                    status.Check(true);
+                    Marshal.FreeHGlobal(protoHandle);
+                }
 
-                    var c_op = c_api.TF_FinishOperation(op_desc, status);
+                var c_op = c_api.TF_FinishOperation(op_desc, status);
 
-                    status.Check(true);
+                status.Check(true);
 
-                    return c_op;
-                }
+                return c_op;
             }
         }
 

src/TensorFlowNET.Core/tensorflow.cs

@@ -22,7 +22,6 @@ using System.Runtime.InteropServices;
 using System.Threading;
 using Tensorflow.Eager;
 using Tensorflow.Gradients;
-using static Tensorflow.Binding;
 
 namespace Tensorflow
 {
@@ -42,11 +41,12 @@ namespace Tensorflow
 
         public delegate Tensor[] BackwardFunction(Tensor[] grads, long[] unneeded_gradients);
 
+        public Status status = new Status();
         public OpDefLibrary _op_def_lib = new OpDefLibrary();
+        public Context context = new Context(new ContextOptions(), new Status());
         public Execute _execute = new Execute();
         public IEagerRunner Runner = new EagerRunner();
-        public Context context = new Context(new ContextOptions(), new Status());
-
+        
         public tensorflow()
         {
             enable_eager_execution();

test/TensorFlowNET.UnitTest/ConstantTest.cs

@@ -96,14 +96,14 @@ namespace TensorFlowNET.UnitTest.Basics
         public void ZerosConst()
         {
             // small size
-            var tensor = tf.zeros(new Shape(3, 2), tf.int32, "small");
+            var tensor = tf.zeros((3, 2), tf.int32, "small");
 
             Assert.AreEqual(tensor.shape[0], 3);
             Assert.AreEqual(tensor.shape[1], 2);
             Assert.IsTrue(Enumerable.SequenceEqual(new int[] { 0, 0, 0, 0, 0, 0 }, tensor.numpy().ToArray<int>()));
 
             // big size
-            tensor = tf.zeros(new Shape(200, 100), tf.int32, "big");
+            tensor = tf.zeros((200, 100), tf.int32, "big");
 
             Assert.AreEqual(tensor.shape[0], 200);
             Assert.AreEqual(tensor.shape[1], 100);

test/TensorFlowNET.UnitTest/NativeAPI/Eager/GradientEagerTest.cs

@@ -35,7 +35,26 @@ namespace TensorFlowNET.UnitTest.Gradient
             var dz_dx = tape.gradient(z, x);
 
             var expected = new float[] { 8.0f, 8.0f, 8.0f, 8.0f };
-            Assert.IsTrue(Enumerable.SequenceEqual(dz_dx.numpy().ToArray<float>(), expected));
+            Assert.IsTrue(Enumerable.SequenceEqual(dz_dx.ToArray<float>(), expected));
+        }
+
+        [TestMethod]
+        public void PersistentTape()
+        {
+            var x = tf.ones((2, 2));
+            using var tape = tf.GradientTape(persistent: true);
+            tape.watch(x);
+            var y = tf.reduce_sum(x);
+            var z = tf.multiply(y, y);
+            var dz_dx = tape.gradient(z, x);
+
+            var expected = new float[] { 8.0f, 8.0f, 8.0f, 8.0f };
+            Assert.IsTrue(Enumerable.SequenceEqual(dz_dx.ToArray<float>(), expected));
+
+            var dz_dy = tape.gradient(z, y);
+
+            expected = new float[] { 8.0f, 8.0f, 8.0f, 8.0f };
+            Assert.IsTrue(Enumerable.SequenceEqual(dz_dx.ToArray<float>(), expected));
         }
     }
 }