Error compiling example jpegNPP

Accelerated Computing CUDA CUDA Setup and Installation
1

Hi,

I installed 5.5 on Ubuntu 12.04 with a GeForce GTX 670.

After the installation I tried to compile the examples with ‘make’, which works w/o problem till jpegNPP where it stops with an error.

jpegNPP.cpp:689:5: error: cannot convert ‘Npp8u** {aka unsigned char**}’ to ‘NppiEncodeHuffmanSpec**’ for argument ‘10’ to ‘NppStatus nppiEncodeHuffmanScan_JPEG_8u16s_P3R(Npp16s**, Npp32s*, Npp32s, Npp32s, Npp32s, Npp32s, Npp32s, Npp8u*, Npp32s*, NppiEncodeHuffmanSpec**, NppiEncodeHuffmanSpec**, NppiSize*, Npp8u*)’

What could be the source of the error? It looks like a type error in the source code, but that seems unrealistic given

Here the full error message:

make[1]: Entering directory `/home/me/cuda5.5/NVIDIA_CUDA-5.5_Samples/7_CUDALibraries/jpegNPP'
/usr/local/cuda-5.5/bin/nvcc -ccbin g++ -I../../common/inc -I../common/UtilNPP -I../common/FreeImage/include  -m64     -o jpegNPP.o -c jpegNPP.cpp
jpegNPP.cpp: In function ‘int main(int, char**)’:
jpegNPP.cpp:568:13: error: cannot convert ‘Npp8u** {aka unsigned char**}’ to ‘NppiDecodeHuffmanSpec**’ for argument ‘10’ to ‘NppStatus nppiDecodeHuffmanScanHost_JPEG_8u16s_P3R(const Npp8u*, Npp32s, Npp32s, Npp32s, Npp32s, Npp32s, Npp32s, Npp16s**, Npp32s*, NppiDecodeHuffmanSpec**, NppiDecodeHuffmanSpec**, NppiSize*)’
jpegNPP.cpp:689:5: error: cannot convert ‘Npp8u** {aka unsigned char**}’ to ‘NppiEncodeHuffmanSpec**’ for argument ‘10’ to ‘NppStatus nppiEncodeHuffmanScan_JPEG_8u16s_P3R(Npp16s**, Npp32s*, Npp32s, Npp32s, Npp32s, Npp32s, Npp32s, Npp8u*, Npp32s*, NppiEncodeHuffmanSpec**, NppiEncodeHuffmanSpec**, NppiSize*, Npp8u*)’
make[1]: *** [jpegNPP.o] Error 1
make[1]: Leaving directory `/home/me/cuda5.5/NVIDIA_CUDA-5.5_Samples/7_CUDALibraries/jpegNPP'
make: *** [7_CUDALibraries/jpegNPP/Makefile.ph_build] Error 2
2

I made make skip jpegNPP by renaming the Makefile in the jpegNPP folder and now he finished with all the other examples without error.

3

Are you running CUDA 5.5 production?

4

Can you please paste your system information here?

uname -a
gcc --version
/usr/local/cuda-5.5/bin/nvcc -V

5

I installed cuda 5.5 via cuda-repo-ubuntu1204_5.5-0_amd64.deb. Using this tutorial:
http://www.r-tutor.com/gpu-computing/cuda-installation/cuda5.5-ubuntu

Here are the system information:

me@serv8:~/cuda5.5$ uname -a
Linux iupr-serv8 3.2.0-48-generic #74-Ubuntu SMP Thu Jun 6 19:43:26 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
me@serv8:~/cuda5.5$ gcc --version
gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
Copyright (C) 2011 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
me@serv8:~/cuda5.5$ /usr/local/cuda-5.5/bin/nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2013 NVIDIA Corporation
Built on Wed_Jul_17_18:36:13_PDT_2013
Cuda compilation tools, release 5.5, V5.5.0
6

Unfortunately, I can’t repro your problem on my Ubuntu 12.04 system. My steps listed below.

/usr/bin/nvidia-uninstall
cd /tmp; wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1204/x86_64/cuda-repo-ubuntu1204_5.5-0_amd64.deb
sudo dpkg -i cuda-repo-ubuntu1204_5.5-0_amd64.deb
sudo apt-get update
sudo apt-get install cuda
/usr/local/cuda-5.5/bin/cuda-install-samples-5.5.sh ~
cd ~/NVIDIA_CUDA-5.5_Samples/7_CUDALibraries/jpegNPP
make

Could you please try command below?
g++ -c -x c++ -D__NVCC__ -I"…/…/common/inc" -I"…/common/UtilNPP" -I"…/common/FreeImage/include" “-I/usr/local/cuda-5.5/bin/…/targets/x86_64-linux/include” -m64 -o “jpegNPP.o” “jpegNPP.cpp” -v

7

:/ I don’t know, what changed, but now it works…

Thanks ryluo!