Skip to content

Conversation

@QiJune
Copy link
Member

@QiJune QiJune commented Aug 31, 2017

Fix #3789

@QiJune QiJune requested review from pkuyym and qingqing01 August 31, 2017 10:18
Copy link
Contributor

@pkuyym pkuyym left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

boost::get<platform::GPUPlace>(src_place), src_ptr, size, 0);
}
PADDLE_ENFORCE(cudaStreamSynchronize(0),
"cudaStreamSynchronize failed in Tensor CopyFrom");
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's better to move cudaStreamSynchronize to pybind. And it's not good to use 0 as cuda stream in the memory::Copy interface.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, I will fix it later. And I set up a issue #3796 , try to add a DeviceContextManager to manage CUDA streams appropriately.

@QiJune QiJune merged commit c2edd2d into PaddlePaddle:develop Sep 1, 2017
heavengate pushed a commit to heavengate/Paddle that referenced this pull request Aug 16, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants