3df79f403e
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/15125 I realized that it is really bad juju if you fake a StreamId out of thin air, because in general this isn't going to work. So, make the constructor a lot scarier. Most "faking StreamId out of thin air" happens because someone just wants to put something on the default stream. Reviewed By: dzhulgakov Differential Revision: D13432800 fbshipit-source-id: a86991d6fc1d8aa4e54e8175e5f06f90856238e6 |
||
|---|---|---|
| .. | ||
| DeviceGuardImplInterface.cpp | ||
| DeviceGuardImplInterface.h | ||
| FakeGuardImpl.h | ||
| InlineDeviceGuard.h | ||
| InlineStreamGuard.h | ||
| README.md | ||
| VirtualGuardImpl.h | ||
c10/impl provides headers for functionality that is only needed in very specific use-cases (e.g., you are defining a new device type), which are generally only needed by C10 or PyTorch code. If you are an ordinary end-user, you should not use headers in this folder. We permanently give NO backwards-compatibility guarantees for implementations in this folder.
Compare with c10/util, which provides functionality that is not directly related to being a deep learning library (e.g., C++20 polyfills), but may still be generally useful and visible to users.
(We don't call this c10/detail, because the detail namespace convention is for header private details. However, c10::impl may be utilized from external headers; it simply indicates that the functionality is not for end users.)