[llvm-dev] Default Location of CUDA headers in Windows and macOS

[SANJAY SRIVALLABH SINGAPURAM via llvm-dev]

Here are the Default locations,
(# is a placeholder for a number ; #.# is the CUDA version number )
Windows : %ProgramFile%\NVIDI GPU Computing Toolkit\CUDA\v#.#
Linux : /usr/local/cuda-#.# ( /usr/local/cuda is a symlink to the latest
version )
macOS : /Developer/NVIDIA/CUDA-#.#

On Mon, May 22, 2017 at 6:02 PM Michael Kruse <llvmdev at meinersbur.de> wrote:

> Never hardcode paths on Windows, they can change per system
> configuration, Windows language or the user might change them to any
> arbitrary location. Look them up using the APIs for this or from the
> registry.
>
It's new to me that the system language can affect paths.

>
> In this case, look for the environment variables "CUDA_PATH" and
> "CUDA_PATH_V8_0" (or some different version) contains the path.
>
Thank You Michael !

>
> Michael
>
>
> 2017-05-22 7:06 GMT+02:00 SANJAY SRIVALLABH SINGAPURAM via llvm-dev
> <llvm-dev at lists.llvm.org>:
> > The paths I found for Windows were dependent on the version of the SDK,
> e.g.
> > C:\Program Files\NVIDIA GPU
> > Computing Toolkit\CUDA\v8.0 or \v7.5. is there a path that's independent
> of
> > the version  ? (/usr/local/cuda for Linux)
> >
> > On Mon 22 May, 2017, 9:58 AM SANJAY SRIVALLABH SINGAPURAM,
> > <llvmresch_int01 at iith.ac.in> wrote:
> >>
> >> Hello,
> >>
> >> Can anyone help me with the default installation locations of CUDA
> headers
> >> in Windows and macOS ? e.g. /usr/local/cuda/include is the default for
> >> Linux.
> >>
> >> Thanks,
> >> Sanjay
> >
> >
> > _______________________________________________
> > LLVM Developers mailing list
> > llvm-dev at lists.llvm.org
> > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170523/f6980440/attachment.html>