From 6fc3e8ec84af9e9fa2f191cbcc402de080962a67 Mon Sep 17 00:00:00 2001
From: danleifeng <52735331+danleifeng@users.noreply.github.com>
Date: Wed, 20 Nov 2019 19:45:20 +0800
Subject: [PATCH] edit elementwise_mul doublegrad inplace (#21245)

---
 .../elementwise/elementwise_div_op.cc         |  2 +-
 .../elementwise/elementwise_div_op.h          |  2 -
 .../elementwise/elementwise_mul_op.cc         |  2 +-
 .../elementwise/elementwise_mul_op.h          | 58 ++++++++++++-------
 4 files changed, 39 insertions(+), 25 deletions(-)

diff --git a/paddle/fluid/operators/elementwise/elementwise_div_op.cc b/paddle/fluid/operators/elementwise/elementwise_div_op.cc
index 0343276b755..115b5b9ef20 100644
--- a/paddle/fluid/operators/elementwise/elementwise_div_op.cc
+++ b/paddle/fluid/operators/elementwise/elementwise_div_op.cc
@@ -128,7 +128,7 @@ REGISTER_OPERATOR(
     ops::ElementwiseDivDoubleGradMaker<paddle::imperative::OpBase>);
 
 REGISTER_OPERATOR(elementwise_div_grad_grad, ops::ElementwiseDivOpDoubleGrad,
-                  ops::ElementwiseDivDoubleGradOpInplace);
+                  ops::ElementwiseDoubleGradOpInplace);
 
 REGISTER_OP_CPU_KERNEL(
     elementwise_div,
diff --git a/paddle/fluid/operators/elementwise/elementwise_div_op.h b/paddle/fluid/operators/elementwise/elementwise_div_op.h
index 8bb1dc4708f..53793571db6 100644
--- a/paddle/fluid/operators/elementwise/elementwise_div_op.h
+++ b/paddle/fluid/operators/elementwise/elementwise_div_op.h
@@ -246,7 +246,5 @@ class ElementwiseDivDoubleGradKernel : public framework::OpKernel<T> {
   }
 };
 
-DECLARE_INPLACE_OP_INFERER(ElementwiseDivDoubleGradOpInplace, {"DDX", "DDOut"});
-
 }  // namespace operators
 }  // namespace paddle
diff --git a/paddle/fluid/operators/elementwise/elementwise_mul_op.cc b/paddle/fluid/operators/elementwise/elementwise_mul_op.cc
index 87dc15fadcc..aaa6cfe0346 100644
--- a/paddle/fluid/operators/elementwise/elementwise_mul_op.cc
+++ b/paddle/fluid/operators/elementwise/elementwise_mul_op.cc
@@ -127,7 +127,7 @@ REGISTER_OPERATOR(
     ops::ElementwiseMulDoubleGradMaker<paddle::imperative::OpBase>);
 
 REGISTER_OPERATOR(elementwise_mul_grad_grad, ops::ElementwiseOpDoubleGrad,
-                  ops::ElementwiseMulDoubleGradOpInplace);
+                  ops::ElementwiseDoubleGradOpInplace);
 
 REGISTER_OP_CPU_KERNEL(
     elementwise_mul,
diff --git a/paddle/fluid/operators/elementwise/elementwise_mul_op.h b/paddle/fluid/operators/elementwise/elementwise_mul_op.h
index cc3226135e3..502da88cf04 100644
--- a/paddle/fluid/operators/elementwise/elementwise_mul_op.h
+++ b/paddle/fluid/operators/elementwise/elementwise_mul_op.h
@@ -172,34 +172,50 @@ class ElementwiseMulDoubleGradKernel : public framework::OpKernel<T> {
     // (2) dy = dout * ddx
     // (3) ddout = ddx * y
     // (4) ddout = ddout + dx
-    // (5) dx = dout *ddy
+    // (5) dx = dout * ddy
     if (ddout) {
-      // use dx to save memory, other than alloc tmp tensor
-      Tensor* ddout_tmp = dx;
-
-      default_elementwise_mul<DeviceContext, T>(ctx, x, &ddy_safe, ddout_tmp);
       int axis = ctx.Attr<int>("axis");
-      // NOTE: in the following ElemwiseGradCompute, for the
-      // first output tensor is nullptr, the branch to calculate first
-      // output tensor will not be activated, DivGradDx function will not
-      // be called and can be ignored, the first branch has little effect
-      // on running speed.
-      ElemwiseGradCompute<DeviceContext, T, MulGradDX<T>, MulGradDY<T>>(
-          ctx, ddx_safe, ddy_safe, *dout, *dout, axis, nullptr, dy,
-          MulGradDX<T>(), MulGradDY<T>());
-      default_elementwise_mul<DeviceContext, T>(ctx, &ddx_safe, y, ddout);
-
       auto& place =
           *ctx.template device_context<DeviceContext>().eigen_device();
-      auto ddout_t = framework::EigenVector<T>::Flatten(*ddout);
-      auto ddout_tmp_t = framework::EigenVector<T>::Flatten(*ddout_tmp);
-      ddout_t.device(place) = ddout_t + ddout_tmp_t;
-      default_elementwise_mul<DeviceContext, T>(ctx, dout, &ddy_safe, dx);
+      // size(ddout) > size(ddx), ddout can't use memory of ddx using inplace
+      if (ddout->numel() > ddx->numel()) {
+        ElemwiseGradCompute<DeviceContext, T, MulGradDX<T>, MulGradDY<T>>(
+            ctx, ddx_safe, ddy_safe, *dout, *dout, axis, dx, dy, MulGradDX<T>(),
+            MulGradDY<T>());
+
+        Tensor ddout_tmp;
+        ddout_tmp.mutable_data<T>(ddout->dims(), ctx.GetPlace());
+
+        default_elementwise_mul<DeviceContext, T>(ctx, y, &ddx_safe, ddout);
+        default_elementwise_mul<DeviceContext, T>(ctx, &ddy_safe, x,
+                                                  &ddout_tmp);
+
+        auto ddout_t = framework::EigenVector<T>::Flatten(*ddout);
+        auto ddout_tmp_t = framework::EigenVector<T>::Flatten(ddout_tmp);
+        ddout_t.device(place) = ddout_t + ddout_tmp_t;
+      } else {
+        // use dx to save memory, other than alloc tmp tensor
+        Tensor* ddout_tmp = dx;
+
+        default_elementwise_mul<DeviceContext, T>(ctx, x, &ddy_safe, ddout_tmp);
+        // NOTE: in the following ElemwiseGradCompute, for the
+        // first output tensor is nullptr, the branch to calculate first
+        // output tensor will not be activated, DivGradDx function will not
+        // be called and can be ignored, the first branch has little effect
+        // on running speed.
+        ElemwiseGradCompute<DeviceContext, T, MulGradDX<T>, MulGradDY<T>>(
+            ctx, ddx_safe, ddy_safe, *dout, *dout, axis, nullptr, dy,
+            MulGradDX<T>(), MulGradDY<T>());
+        default_elementwise_mul<DeviceContext, T>(ctx, &ddx_safe, y, ddout);
+
+        auto ddout_t = framework::EigenVector<T>::Flatten(*ddout);
+        auto ddout_tmp_t = framework::EigenVector<T>::Flatten(*ddout_tmp);
+        ddout_t.device(place) = ddout_t + ddout_tmp_t;
+        default_elementwise_mul<DeviceContext, T>(ctx, dout, &ddy_safe, dx);
+      }
     }
   }
 };
 
-DECLARE_INPLACE_OP_INFERER(ElementwiseMulDoubleGradOpInplace, {"DDX", "DDOut"});
-
 }  // namespace operators
 }  // namespace paddle
-- 
GitLab