cutil_inline.h

Accelerated Computing CUDA CUDA Programming and Performance
1

Hey,

When I try to follow the instructions as Create your own SDK in the programming guide and copy the sample template to my home directory and try to do a make, I get the following error: cannot find cutil_inline.h

I have set the PATHS as mentioned. Please Help.

2

Do you have ???/nvidia_cuda_sdk/common/inc in your path? That’s what I needed.

Malcolm

3

I think that even you doing this, when you compile the code, the executable goes to NVIDIA_GPU_Computing_SDK/C/linux/release/. I solve this editing file common.mk and writing my own makefile to compile the code.

4

Luis, you are corect. My suggestion was for SDK 2.2. For SDK 2.3 you would use “NVIDIA_GPU_Computing_SDK/C/common/common.mk”, I believe.

Malcolm

5

If anyone want the file I wrote just email me. I modified the file commons.mk in order to use the home folder for a project anywhere. The release, debug or emudebug binaries are generated inside the project folder.

Hope this will help you.

6

Hi,
anyone can help? This is a legacy question. I understand the SDK is now part of the Toolkit, but after installing CUDA 6.5, there is no common.mk and no cutil_inline.h anymore. Can anyone confirm?
Regards

7

I confirm. The cutil stuff was removed.

There are numerous questions here on this board as well as all over the web, describing various approaches to dealing with this or working around it.

for example:

[url]https://devtalk.nvidia.com/default/topic/827637/cuda-programming-and-performance/mars-mapreduce-framework-amp-cuda-toolkit-7-0/[/url]

and:

[url]https://devtalk.nvidia.com/default/topic/758475/fatal-error-cutil-h-no-such-file-or-directory-include-lt-cutil-h-gt-compilation-terminated/?offset=1[/url]

8

Thanks, I guess I need to change something here then.