Commit Graph

5 Commits

Author SHA1 Message Date
Luca Wehrstedt
1ac05cfe01 Remove DataPtr extractor from CUDAFuture (#48840)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/48840

The CUDAFuture class needs to inspect the values it contains in order to extract its tensors (in fact, the DataPtrs backing those). These are needed first to determine what CUDA devices back those tensors, so that an event for each such device can be recorded; and later to record these DataPtrs with the CUDA caching allocator if they are used in other streams.

This became complicated when Python was added to the mix, because to inspect a Python object we need to acquire the GIL, but we couldn't do so from code that was supposed to also work in C++-only mode. The solution was for users to provide a custom way to extract DataPtrs, so that the PythonFutureWrapper could install such a custom Python-aware one. This was the DataPtr extractor.

In https://github.com/pytorch/pytorch/pull/48502 a different suggestion was proposed. At its root, it consists in adding support for IValues of type PyObject to the visit() and getSubValues() methods. In order to deal with the GIL, we do this through a virtual method: PyObjectHolder, which is the base class, is available also in C++-only mode, and thus defines this method but leaves it unimplemented; ConcretePyObjectHolder, which is the subclass, is only included in Python mode, and thus it can implement that method, acquire the GIL, and do what it's supposed to.

In my opinion, this approach is just brilliant! Thank wanchaol for proposing it! It hides the complexity of dealing with Python inside getSubValues(), where it can be done properly, thus simplifying enormously the CUDAFuture and the PythonFutureWrapper classes.
ghstack-source-id: 118704935

Test Plan: Unit tests

Reviewed By: wanchaol

Differential Revision: D25334355

fbshipit-source-id: 3f1d3bf6e6e8505a114c877fb9a6fcc3f68d91d3
2020-12-19 11:03:45 -08:00
Shen Li
2e9d6d99be Explicitly decref py::object in ConcretePyObjectHolder and PythonFunctionGuard (#38364)
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/38364

Test Plan: Imported from OSS

Differential Revision: D21537611

Pulled By: mrshenli

fbshipit-source-id: e22d1f1360cf71bec526841b5014013b11316f8d
2020-05-12 20:55:53 -07:00
Shen Li
ee1ddcef8d Acquire GIL when constructing/destructing ConcretePyObjectHolder (#37870)
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/37870

Test Plan: Imported from OSS

Differential Revision: D21410785

fbshipit-source-id: 374d5f40fbdfec98262aa4c84ec4ccdc40fb2ac1
2020-05-07 07:37:39 -07:00
Meghan Lele
6384c2d81b [JIT] clang-format JIT code (#35115)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/35115

This commit runs the newly added tools/clang_format.py on the JIT
codebase and includes all of the formatting changes thus produced.

Testing:
Ran the script, CI.

Test Plan: Imported from OSS

Reviewed By: eellison

Differential Revision: D20568523

Pulled By: SplitInfinity

fbshipit-source-id: e09bdb982ccf090eecfb7c7b461b8d0681eef82b
2020-03-26 11:24:51 -07:00
Michael Suo
dbe850af5b [jit] do the code reorg (#33851)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/33851

Rationale and context described in #33828.

Script to reproduce the move:
https://gist.github.com/suo/16cbefaaeb67ca5a7c6caffd49b7f6e9
ghstack-source-id: 99079645

Test Plan: Make sure CI passes

Reviewed By: jamesr66a

Differential Revision: D20133869

fbshipit-source-id: 390e9241a9c85366d9005c492ac31f10aa96488e
2020-02-27 13:02:51 -08:00