Windows 10 build failed

Open
#846 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
25/100
Issue type
Bug
Clarity
Needs clarification
Activity status
Stale
Tech stack
cpp, python, pytorch

Research direction

Start with csrc/multi_tensor_scale_kernel.cu at lines 69 and 97, then reproduce the Windows 10 build using the reported PyTorch, CUDA toolkit, GPU driver, and NVIDIA nvcc versions. Trace the device-code compilation errors involving isfinitec10::Half; done means the extension builds successfully without these errors on the reported environment.

Written by the indexing model from the issue text.

Description

csrc/multi_tensor_scale_kernel.cu(69): error: calling a host function("isfinite< ::c10::Half> ") from a device function("ScaleFunctor< ::c10::Half, float> ::operator ()") is not allowed

csrc/multi_tensor_scale_kernel.cu(69): error: identifier "isfinite< ::c10::Half> " is undefined in device code

csrc/multi_tensor_scale_kernel.cu(97): error: calling a host function("isfinite< ::c10::Half> ") from a device function("ScaleFunctor< ::c10::Half, float> ::operator ()") is not allowed

csrc/multi_tensor_scale_kernel.cu(97): error: identifier "isfinite< ::c10::Half> " is undefined in device code

csrc/multi_tensor_scale_kernel.cu(69): error: calling a host function("isfinite< ::c10::Half> ") from a device function("ScaleFunctor< ::c10::Half, ::c10::Half> ::operator ()") is not allowed

csrc/multi_tensor_scale_kernel.cu(69): error: identifier "isfinite< ::c10::Half> " is undefined in device code

csrc/multi_tensor_scale_kernel.cu(97): error: calling a host function("isfinite< ::c10::Half> ") from a device function("ScaleFunctor< ::c10::Half, ::c10::Half> ::operator ()") is not allowed

csrc/multi_tensor_scale_kernel.cu(97): error: identifier "isfinite< ::c10::Half> " is undefined in device code

8 errors detected in the compilation of "C:/Users/morri/AppData/Local/Temp/tmpxft_000015ec_00000000-10_multi_tensor_scale_kernel.cpp1.ii".
error: command 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\bin\nvcc.exe' failed with exit status 1

====================================
OS: Windows 10
GPU: NVIDIA-SMI 445.87 Driver Version: 445.87 CUDA Version: 11.0
Packages: pytorch 1.0.0
cudatoolkit 10.0.130

Dominant language
Python
Stars
9k
Forks
1.5k
Avg merge
2d 4h
Merged PRs (30d)
3

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from NVIDIA/apex

All issues in NVIDIA/apex

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.