[MKL-DNN] Pool & LRN Grad Ops NHWC support (#21747)

paddle/fluid/framework/data_layout_transform.cc

@@ -185,10 +185,8 @@ void innerTransDataLayoutFromMKLDNN(DataLayout in_layout, DataLayout out_layout,
   }
   // For exepected NHWC data format we need to reshape the Output tensor
   // As MKL-DNN description was in NCHW and paddle is expecting NHWC
-  if (out_layout == DataLayout::kNHWC) {
-    std::rotate(out_tz.begin() + 1, out_tz.begin() + 2, out_tz.end());
-    out->Resize(framework::make_ddim(out_tz));
-  }
+  platform::MatchShapeToLayout(out, in_layout, out_layout);
+
   out->set_layout(out_layout);
   // reset format since the out tensor will be feed to non-MKLDNN OPkernel
   out->set_format(MKLDNNMemoryFormat::undef);

paddle/fluid/framework/data_transform.cc

@@ -58,13 +58,8 @@ void TransformData(const OpKernelType &expected_kernel_type,
         out.ShareDataWith(input_tensor);
         // For NHWC data we need reshape of tensors as MKL-DNN
         // is expecting NHWC dims description order
-        if (lin == DataLayout::kNHWC) {
-          auto nchw_dims = paddle::framework::vectorize<int>(out.dims());
-          std::rotate(nchw_dims.begin() + 1, nchw_dims.end() - 1,
-                      nchw_dims.end());
-          out.Resize(framework::make_ddim(nchw_dims));
-          paddle::platform::set_cur_paddle_data_layout(lin);
-        }
+        platform::MatchShapeToLayout(&out, lin, lout);
+        paddle::platform::set_cur_paddle_data_layout(lin);
         out.set_layout(DataLayout::kMKLDNN);
         out.set_format(out_format);
       } else {

paddle/fluid/framework/data_transform.h

@@ -39,6 +39,5 @@ void TransformData(const OpKernelType &expected_kernel_type,
  */
 void SetTensorToVariable(const Variable &in_var, const Tensor &tensor,
                          Variable *out_var);
-
 }  // namespace framework
 }  // namespace paddle

paddle/fluid/framework/operator.cc

@@ -33,6 +33,10 @@ limitations under the License. */
 #include "paddle/fluid/framework/var_type.h"
 #include "paddle/fluid/platform/profiler.h"
 
+#ifdef PADDLE_WITH_MKLDNN
+#include "paddle/fluid/platform/mkldnn_helper.h"
+#endif
+
 DECLARE_bool(benchmark);
 DECLARE_bool(check_nan_inf);
 DECLARE_bool(enable_unused_var_check);
@@ -1102,11 +1106,8 @@ Scope* OperatorWithKernel::PrepareData(
   }
 
   for (auto& var_name_item : Inputs()) {
-    if (no_buffer_ins && no_buffer_ins->count(var_name_item.first) > 0) {
-      VLOG(7) << "Skip scanning input " << var_name_item.first
-              << " in Operator " << type_;
-      continue;
-    }
+    bool should_skip_input =
+        no_buffer_ins && no_buffer_ins->count(var_name_item.first) > 0;
 
     std::vector<Variable*>& input_vars = ctx->inputs[var_name_item.first];
 
@@ -1120,6 +1121,44 @@ Scope* OperatorWithKernel::PrepareData(
       }
 
       auto* tensor_in = GetLoDTensorOrSelectedRowsValueFromVar(*var);
+
+      // When no_buffer_ins then checking of Tensor::holder_ is
+      // not a thread safe. And for infershape scenario checks
+      // to be omitted are not really needed
+      if (should_skip_input == true) {
+#ifdef PADDLE_WITH_MKLDNN
+        // Var without buffer may be needed
+        // for some situation like InferShape().
+        // In this situation We cannot skip Var analysis, as
+        // MKL-DNN shape of Var may differ from kNHWC Var
+        // In such situation corressponding resized Var
+        // has to be created and registered
+        if ((tensor_in->layout() == DataLayout::kMKLDNN) &&
+            (var->IsType<LoDTensor>() == true) &&
+            (expected_kernel_key.data_layout_ != DataLayout::kMKLDNN) &&
+            (paddle::platform::get_cur_paddle_data_layout() ==
+             DataLayout::kNHWC)) {
+          // Mixed execution : MKL-DNN and GPU is not supported!
+          if (!new_scope) {
+            new_scope = &scope.NewScope();
+          }
+          auto* trans_var = new_scope->Var(var_name);
+          input_vars[i] = trans_var;
+          auto out = trans_var->GetMutable<LoDTensor>();
+          out->Resize(tensor_in->dims());
+          platform::MatchShapeToLayout(out, tensor_in->layout(),
+                                       DataLayout::kNHWC);
+          VLOG(7) << "Created reshaped dummy input based on MKL-DNN Tensor , "
+                     "but kNHWC layout"
+                  << var_name_item.first << " in Operator " << type_;
+        } else {
+          VLOG(7) << "Skip scanning input " << var_name_item.first
+                  << " in Operator " << type_;
+        }
+#endif
+        continue;
+      }
+
       if (!tensor_in->IsInitialized()) {
         continue;
       }
@@ -1143,14 +1182,17 @@ Scope* OperatorWithKernel::PrepareData(
       // In the inference scenerio, the scopes will be reused across the
       // batches, so the `new_scope` here will result in GPU memroy explosion
       // over the  running of operators.
-      // We use a thread_local cache to fix that issue, the key in the cache is
+      // We use a thread_local cache to fix that issue, the key in the cache
+      // is
       // the combination of the `scope` argument, from_kernel_type,
       // target_kernel_type.
       // Have a discussion with @Superjomn or the inference developers if some
       // changes on this logic for this macro might not tested on the other
       // scenerios.
-      // If this op is not called by an Executor or ParallelExecutor, it should
-      // called by a NaiveExecutor, the NaiveExecutor will cache the scopes and
+      // If this op is not called by an Executor or ParallelExecutor, it
+      // should
+      // called by a NaiveExecutor, the NaiveExecutor will cache the scopes
+      // and
       // variables, that behavior a lot different.
       //
       // To solve issue #15032, have a discussion with @Luotao for cpu
@@ -1174,15 +1216,14 @@ Scope* OperatorWithKernel::PrepareData(
       // we will create a new cpu tensor in new scope.
       // However, if enable_cache_runtime_context_, we get the cpu tensor each
       // time, not the gpu tensor.
-      // Thus, we set pre_scope_ = nullptr to trigger `new RuntimeContext()` in
+      // Thus, we set pre_scope_ = nullptr to trigger `new RuntimeContext()`
+      // in
       // RunImpl().
       if (enable_cache_runtime_context_) {
         pre_scope_ = nullptr;
       }
-
       auto* trans_var = new_scope->Var(var_name);
       input_vars[i] = trans_var;
-
       Tensor out;
       TransformData(expected_kernel_key, kernel_type_for_var, *tensor_in, &out);
       SetTensorToVariable(*var, out, trans_var);

paddle/fluid/operators/lrn_op.cc

@@ -334,12 +334,6 @@ class LRNOpGrad : public framework::OperatorWithKernel {
 #ifdef PADDLE_WITH_MKLDNN
     if (library_ == framework::LibraryType::kPlain &&
         platform::CanMKLDNNBeUsed(ctx)) {
-      // TODO(jczaja): Add support for NHWC
-      const std::string data_format = ctx.Attr<std::string>("data_format");
-      PADDLE_ENFORCE_NE(
-          data_format, "NHWC",
-          platform::errors::Unimplemented(
-              "LRN MKLDNN grad does not support NHWC data format yet"));
       library_ = framework::LibraryType::kMKLDNN;
       layout_ = framework::DataLayout::kMKLDNN;
     }
@@ -348,6 +342,28 @@ class LRNOpGrad : public framework::OperatorWithKernel {
         OperatorWithKernel::IndicateVarDataType(ctx, "X"), ctx.GetPlace(),
         layout_, library_);
   }
+
+  framework::OpKernelType GetKernelTypeForVar(
+      const std::string& var_name, const Tensor& tensor,
+      const framework::OpKernelType& expected_kernel_type) const override {
+#ifdef PADDLE_WITH_MKLDNN
+    if ((expected_kernel_type.data_layout_ == framework::DataLayout::kMKLDNN) &&
+        (tensor.layout() != framework::DataLayout::kMKLDNN)) {
+      auto attrs = Attrs();
+      auto ar = paddle::framework::AttrReader(attrs);
+      const std::string data_format = ar.Get<std::string>("data_format");
+      auto dl = framework::StringToDataLayout(data_format);
+      // Some models may have intentionally set "AnyLayout" for lrn
+      // op. Treat this as NCHW (default data_format value)
+      if (dl != framework::DataLayout::kAnyLayout) {
+        return framework::OpKernelType(expected_kernel_type.data_type_,
+                                       tensor.place(), dl);
+      }
+    }
+#endif
+    return framework::OpKernelType(expected_kernel_type.data_type_,
+                                   tensor.place(), tensor.layout());
+  }
 };
 
 template <typename T>

paddle/fluid/operators/pool_op.cc

@@ -202,12 +202,6 @@ framework::OpKernelType PoolOpGrad::GetExpectedKernelType(
 #ifdef PADDLE_WITH_MKLDNN
   if (library_ == framework::LibraryType::kPlain &&
       platform::CanMKLDNNBeUsed(ctx)) {
-    // TODO(jczaja): Add support for NHWC
-    const std::string data_format = ctx.Attr<std::string>("data_format");
-    PADDLE_ENFORCE_NE(
-        data_format, "NHWC",
-        platform::errors::Unimplemented(
-            "Pool MKLDNN grad does not support NHWC data format yet"));
     library_ = framework::LibraryType::kMKLDNN;
     layout_ = framework::DataLayout::kMKLDNN;
   }
@@ -222,6 +216,24 @@ framework::OpKernelType PoolOpGrad::GetExpectedKernelType(
                                  library_);
 }
 
+framework::OpKernelType PoolOpGrad::GetKernelTypeForVar(
+    const std::string& var_name, const Tensor& tensor,
+    const framework::OpKernelType& expected_kernel_type) const {
+#ifdef PADDLE_WITH_MKLDNN
+  if ((expected_kernel_type.data_layout_ == framework::DataLayout::kMKLDNN) &&
+      (tensor.layout() != framework::DataLayout::kMKLDNN)) {
+    auto attrs = Attrs();
+    auto ar = paddle::framework::AttrReader(attrs);
+    const std::string data_format = ar.Get<std::string>("data_format");
+    return framework::OpKernelType(expected_kernel_type.data_type_,
+                                   tensor.place(),
+                                   framework::StringToDataLayout(data_format));
+  }
+#endif
+  return framework::OpKernelType(expected_kernel_type.data_type_,
+                                 tensor.place(), tensor.layout());
+}
+
 void Pool2dOpMaker::Make() {
   AddInput(
       "X",

paddle/fluid/operators/pool_op.h

@@ -38,7 +38,7 @@ class PoolOp : public framework::OperatorWithKernel {
 
   framework::OpKernelType GetKernelTypeForVar(
       const std::string& var_name, const Tensor& tensor,
-      const framework::OpKernelType& expected_kernel_type) const;
+      const framework::OpKernelType& expected_kernel_type) const override;
 };
 
 class PoolOpGrad : public framework::OperatorWithKernel {
@@ -50,6 +50,10 @@ class PoolOpGrad : public framework::OperatorWithKernel {
  protected:
   framework::OpKernelType GetExpectedKernelType(
       const framework::ExecutionContext& ctx) const override;
+
+  framework::OpKernelType GetKernelTypeForVar(
+      const std::string& var_name, const Tensor& tensor,
+      const framework::OpKernelType& expected_kernel_type) const override;
 };
 
 class Pool2dOpMaker : public framework::OpProtoAndCheckerMaker {

paddle/fluid/platform/mkldnn_helper.h

@@ -71,6 +71,29 @@ tf_pd<Type> MKLDNNBwdPrimitiveDesc(const Engine& e, const Primitive& p,
   return tf_pd<Type>(desc, e, p);
 }
 
+inline void MatchShapeToLayout(framework::Tensor* tensor_in,
+                               framework::DataLayout from,
+                               framework::DataLayout to) {
+  switch (from) {
+    case framework::DataLayout::kMKLDNN:
+      if (to == framework::DataLayout::kNHWC) {
+        auto dims = framework::vectorize<int>(tensor_in->dims());
+        std::rotate(dims.begin() + 1, dims.begin() + 2, dims.end());
+        tensor_in->Resize(framework::make_ddim(dims));
+      }
+      break;
+    case framework::DataLayout::kNHWC:
+      if (to == framework::DataLayout::kMKLDNN) {
+        auto dims = framework::vectorize<int>(tensor_in->dims());
+        std::rotate(dims.begin() + 1, dims.end() - 1, dims.end());
+        tensor_in->Resize(framework::make_ddim(dims));
+      }
+      break;
+    default:
+      break;
+  }
+}
+
 inline mkldnn::memory::desc MKLDNNMemDesc(const std::vector<int64_t>& dims,
                                           mkldnn::memory::data_type data_type,
                                           MKLDNNMemoryFormat format) {

python/paddle/fluid/tests/unittests/mkldnn/test_lrn_mkldnn_op.py

@@ -59,11 +59,6 @@ class TestLRNMKLDNNOpNHWC(TestLRNMKLDNNOp):
     def init_test_case(self):
         self.data_format = 'NHWC'
 
-    #TODO(jczaja): Add grad support
-    def test_check_grad_normal(self):
-        with self.assertRaises(fluid.core_avx.EnforceNotMet):
-            self.check_grad(['X'], 'Out')
-
 
 if __name__ == "__main__":
     unittest.main()

python/paddle/fluid/tests/unittests/mkldnn/test_pool2d_mkldnn_op.py

@@ -157,11 +157,6 @@ class TestAsymPadValidNHWC(TestAsymPadValid):
     def init_shape(self):
         self.shape = [2, 7, 7, 3]
 
-    #TODO(jczaja): Add Grad NHWC support
-    def test_check_grad(self):
-        with self.assertRaises(fluid.core_avx.EnforceNotMet):
-            super(TestAsymPadValidNHWC, self).test_check_grad()
-
 
 if __name__ == '__main__':
     unittest.main()