Issue description

Compiling pytorch on MacOSX 10.13.5 with support for CUDA GeForceGT 750M 2048MB

Code example

Error log
/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu:222:134: warning: self-comparison always evaluates to true [-Wtautological-compare]
switch (memType) { case CUDAHistogramMemoryType::SHARED: (__cudaPushCallConfiguration(grid, block, (CUDAHistogramMemoryType::SHARED == CUDAHistogramMemoryType::SHARED) ? sharedMem : (0), at::globalContext().getCurrentCUDAStream())) ? (void)0 : kernelHistogram1D< output_t, input_t, int64_t, 1, 2, 1, CUDAHistogramMemoryType::SHARED> (aInfo, pInfo, bInfo, binsize, totalElements, getWeightsOp); if (!((cudaGetLastError()) == (cudaSuccess))) { throw Error({func, "/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu", 222}, at::str(at::str("cudaGetLastError() == cudaSuccess", " ASSERT FAILED at ", "/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu", ":", 222, ", please report a bug to PyTorch. ", "kernelHistogram1D failed"))); } ; ; break; case CUDAHistogramMemoryType::MULTI_BLOCK: (__cudaPushCallConfiguration(grid, block, (CUDAHistogramMemoryType::MULTI_BLOCK == CUDAHistogramMemoryType::SHARED) ? sharedMem : (0), at::globalContext().getCurrentCUDAStream())) ? (void)0 : kernelHistogram1D< output_t, input_t, int64_t, 1, 2, 1, CUDAHistogramMemoryType::MULTI_BLOCK> (aInfo, pInfo, bInfo, binsize, totalElements, getWeightsOp); if (!((cudaGetLastError()) == (cudaSuccess))) { throw Error({func, "/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu", 222}, at::str(at::str("cudaGetLastError() == cudaSuccess", " ASSERT FAILED at ", "/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu", ":", 222, ", please report a bug to PyTorch. ", "kernelHistogram1D failed"))); } ; ; break; default: (__cudaPushCallConfiguration(grid, block, (CUDAHistogramMemoryType::GLOBAL == CUDAHistogramMemoryType::SHARED) ? sharedMem : (0), at::globalContext().getCurrentCUDAStream())) ? (void)0 : kernelHistogram1D< output_t, input_t, int64_t, 1, 2, 1, CUDAHistogramMemoryType::GLOBAL> (aInfo, pInfo, bInfo, binsize, totalElements, getWeightsOp); if (!((cudaGetLastError()) == (cudaSuccess))) { throw Error({func, "/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu", 222}, at::str(at::str("cudaGetLastError() == cudaSuccess", " ASSERT FAILED at ", "/Users/josepen/Development/pytorch/aten/src/ATen/native/cuda/SummaryOps.cu", ":", 222, ", please report a bug to PyTorch. ", "kernelHistogram1D failed"))); } ; ; }

System Info

josepens-mbp:pytorch josepen$ python torch/utils/collect_env.py
Collecting environment information...
PyTorch version: 0.5.0a0+290d20b
Is debug build: No
CUDA used to build PyTorch: 9.2

OS: Mac OSX 10.13.5
GCC version: Could not collect
CMake version: version 3.11.1

Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 9.2.64
GPU models and configuration: Could not collect
Nvidia driver version: Could not collect
cuDNN version: Probably one of the following:
/usr/local/cuda/lib/libcudnn.7.dylib
/usr/local/cuda/lib/libcudnn.dylib
/usr/local/cuda/lib/libcudnn_static.a

Versions of relevant libraries:
[pip3] numpy (1.14.5)
[pip3] torch (0.3.1)
[pip3] torchtext (0.2.3)
[pip3] torchvision (0.2.1)
[conda] torch 0.5.0a0+290d20b
[conda] torchtext 0.2.3
[conda] torchvision 0.2.1
josepens-mbp:pytorch josepen$

• PyTorch or Caffe2: pytorch
• How you installed PyTorch (conda, pip, source): I'm compiling from source
• Build command you used (if compiling from source): MACOSX_DEPLOYMENT_TARGET=10.13 CC=clang CXX=clang++ python setup.py install
• OS: MacOSX 10.13.5
• PyTorch version: I cloned the repo this is the hash version
  josepens-mbp:pytorch josepen$ git rev-parse HEAD
  290d20b
• Python version: 3.6
• CUDA/cuDNN version: 9.2.64
• GPU models and configuration: GeForceGT 750M
• GCC version (if compiling from source): gcc --version
  Configured with: --prefix=/Library/Developer/CommandLineTools/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
  Apple LLVM version 9.1.0 (clang-902.0.39.2)
  Target: x86_64-apple-darwin17.6.0
  Thread model: posix
  InstalledDir: /Library/Developer/CommandLineTools/usr/bin
• CMake version: cmake --version
  cmake version 3.11.1
  CMake suite maintained and supported by Kitware (kitware.com/cmake)
• Versions of any other relevant libraries: