justanotheruser opened a new issue #20099:
URL: https://github.com/apache/incubator-mxnet/issues/20099
## Description
mxnet_60 project fails to build
### Error Message
Lots of them but only of two kinds:
``` Code Description Project File Line ```
```
calling a __host__ function("__floorf") from a __global__
function("mxnet::op::adaptiveaveragepool<float> ") is not allowed mxnet_60
D:\clean_version\src\src\operator\contrib\adaptive_avg_pooling.cu 73
```
```
identifier "__floorf" is undefined in device code mxnet_60
D:\clean_version\src\src\operator\contrib\adaptive_avg_pooling.cu 73
```
### Steps to reproduce
1. mkdir build
2. git clone --branch v1.8.x https://github.com/apache/incubator-mxnet.git
src
3. cmake -G "Visual Studio 16 2019" -A x64^
-T cuda=10.2,host=x64^
-DCMAKE_POLICY_DEFAULT_CMP0104=NEW^
-DCMAKE_BUILD_TYPE=Release^
-DUSE_CPP_PACKAGE=1^
-DUSE_CUDA=1^
-DUSE_CUDNN=1^
-DUSE_OPENCV=0^
-DUSE_OPENMP=1^
-DUSE_BLAS=mkl^
-DMKL_ROOT="C:\Program Files
(x86)\IntelSWTools\compilers_and_libraries_2020\windows\mkl"^
-DMXNET_CUDA_ARCH=6.0;6.2;7.0;7.2;7.5^
-DCUDNN_INCLUDE="C:\Program Files\NVIDIA GPU Computing
Toolkit\CUDA\v10.2\include"^
-DCUDNN_LIBRARY="C:\Program Files\NVIDIA GPU Computing
Toolkit\CUDA\v10.2\lib\x64\cudnn.lib"^
-S .\src -B .\build
4. Build solution in VS 2019
## What have you tried to solve it?
1. Updating to latest patch of VS code Microsoft Visual Studio Community 2019
2. Reinstalling CUDA and cuDNN (after VS)
Version 16.9.2
VisualStudio.16.Release/16.9.2+31112.23
Microsoft .NET Framework
Version 4.8.04084
Installed Version: Community
Visual C++ 2019 00435-60000-00000-AA092
Microsoft Visual C++ 2019
NVIDIA CUDA 10.2 Wizards 10.2
NVIDIA Nsight Visual Studio Edition - CUDA support 2019.4.0.19274
3. Building with VS2017
Actually that how I wanted to build it initially but I've got `compiler is
out of heap space in pass 2` error for the same project which I couldn't
resolve (as suggested somewhere here) by settings 'Link Time Code Optimization'
to '/LTCG:PGOptimize' because VS can't find mxnet_60.pgd file. However, I've
successfully built version without CUDA with VS 2019.
## Environment
<summary>Environment Information</summary>
```
(base) C:\Users\User\Desktop>python diagnose.py
----------Python Info----------
Version : 3.8.3
Compiler : MSC v.1916 64 bit (AMD64)
Build : ('default', 'Jul 2 2020 17:30:36')
Arch : ('64bit', 'WindowsPE')
------------Pip Info-----------
Version : 20.1.1
Directory : C:\Users\User\anaconda3\lib\site-packages\pip
----------MXNet Info-----------
An error occured trying to import mxnet.
This is very likely due to missing missing or incompatible library files.
Traceback (most recent call last):
File "diagnose.py", line 96, in check_mxnet
print('Version :', mxnet.__version__)
AttributeError: module 'mxnet' has no attribute '__version__'
----------System Info----------
Platform : Windows-10-10.0.19041-SP0
system : Windows
node : Slepcov-PC
release : 10
version : 10.0.19041
----------Hardware Info----------
machine : AMD64
processor : Intel64 Family 6 Model 165 Stepping 5, GenuineIntel
Name
Intel(R) Core(TM) i7-10700 CPU @ 2.90GHz
----------Network Test----------
Setting timeout: 10
Timing for MXNet: https://github.com/apache/incubator-mxnet, DNS: 0.0338
sec, LOAD: 0.5731 sec.
Timing for Gluon Tutorial(en): http://gluon.mxnet.io, DNS: 0.2435 sec, LOAD:
0.3744 sec.
Error open Gluon Tutorial(cn): https://zh.gluon.ai, <urlopen error [SSL:
CERTIFICATE_VERIFY_FAILED] certificate verify failed: certificate has expired
(_ssl.c:1108)>, DNS finished in 0.18741607666015625 sec.
Timing for FashionMNIST:
https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/dataset/fashion-mnist/train-labels-idx1-ubyte.gz,
DNS: 0.1526 sec, LOAD: 0.7710 sec.
Timing for PYPI: https://pypi.python.org/pypi/pip, DNS: 0.0227 sec, LOAD:
0.7252 sec.
Error open Conda: https://repo.continuum.io/pkgs/free/, HTTP Error 403:
Forbidden, DNS finished in 0.14266681671142578 sec.
----------Environment----------
MXNET_CUDA="D:\Recognition Technologies\AvtoUragan ver
3.8\ExtLibs\MxNet\Win64"
```
</details>
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
