LinuxQuestions.org
Help answer threads with 0 replies.
Go Back   LinuxQuestions.org > Forums > Non-*NIX Forums > Programming
User Name
Password
Programming This forum is for all programming questions.
The question does not have to be directly related to Linux and any language is fair game.

Notices


Reply
  Search this Thread
Old 07-29-2011, 09:51 PM   #1
bxsjc
LQ Newbie
 
Registered: Jul 2011
Posts: 5

Rep: Reputation: Disabled
CUDA on fedora 15


Hi! I came across some problems. I need to conduct parallel computing on CUDA. I input "uname -a"and read these:
"Linux BXSJC 2.6.38.8-35.fc15.i686 #1 SMP Wed Jul 6 14:46:26 UTC 2011 i686 i686 i386 GNU/Linux".
I use CUDAtoolkit 4.0. However, when I input:
cd /home/bxsjc/NVIDIA_GPU_Computing_SDK/C and "make -i ",the computer returns

"/usr/local/cuda/include/host_config.h:82:2: error: #error — unsupported GNU version! gcc 4.5 and up are not supported!
make[1]: [obj/i386/release/clock.cu.o] Error 1 (ignored)
g++: error: obj/i386/release/clock.cu.o: No such file or directory
make[1]: [../../bin/linux/release/clock] Error 1 (ignored)
make[1]: Leaving directory `/home/bxsjc/NVIDIA_GPU_Computing_SDK/C/src/clock'
Finished building all"
I don't know how to tackle this problem. So I directly "cd /home/bxsjc/NVIDIA_GPU_Computing_SDK/C/bin/linux/release "and run those executable files. Some return passed while others return these:
"[bxsjc@BXSJC release]$ ./simpleTextureDrv
[simpleTextureDrv] starting...
> Using CUDA Device [0]: GeForce G210M
> GPU Device has SM 1.2 compute capability
> findModulePath file not found: <simpleTexture_kernel.ptx>
> findModulePath file not found: <simpleTexture_kernel.cubin>
> findModulePath could not find <simpleTexture_kernel> ptx or cubin
[simpleTextureDrv] test results...
FAILED
Press ENTER to exit..."
So, could anyone help me? Any suggestions are appreciated!

Last edited by bxsjc; 08-03-2011 at 11:44 AM.
 
Old 07-30-2011, 02:05 AM   #2
John VV
LQ Muse
 
Registered: Aug 2005
Location: A2 area Mi.
Posts: 17,437

Rep: Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591
have you read the nvidia web site ? and the nvidia dev forum ?
http://developer.nvidia.com/category/zone/cuda-zone
from
http://developer.nvidia.com/cuda-toolkit-40#Linux
fedora 13 is the last supported for the cuda4 toolkit
it will require gcc 4.4 or 4.1
build it or it might be in the fedora repos
i do know that the old 3 built in gcc 4.1

right now i am DL'ing
h?p://developer.download.nvidia.com/compute/cuda/4_0/sdk/gpucomputingsdk_4.0.17_linux.run
BUT
i am running gcc4.5( or 4.3,4.1,or 3.4) and NOT NOT gcc 4.6
i do not think you can build the sdk using gcc4.6
4.6 is way WAY WAY to new .

did you read and fallow the "CUDA_SDK_Release_Notes.txt"
NVIDIA_GPU_Computing_SDK/doc/CUDA_SDK_Release_Notes.txt
Code:
--------------------------------------------------------------------------------
I. (b)  Linux Installation Instructions
--------------------------------------------------------------------------------
[Part 1/2]

	Note: The default installation folder <SDK_INSTALL_PATH> is "~/NVIDIA_GPU_Computing_SDK"

	For more detailed instructions, see section II below.

	0. Install the NVIDIA Linux display driver by executing the file

	   a. For 32-bit linux distributions use:
		  cudadriver_4.0_linux_32_270.xx.run

	   b. For 64-bit linux distributions use:
		  cudadriver_4.0_linux_64_270.xx.run
	  
	   For information on installing NVIDIA Linux display drivers, please refer to 
	   the NVIDIA Accelerated Linux Driver Set README and Installation Guide:
	   http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9755/README/index.html

	1. Install version 4.0 Release of the NVIDIA Toolkit by executing the file
	   cudatoolkit_4.0_linux_*.run where * corresponds to your Linux distribution

	   Add the CUDA binaries and lib path to your PATH and LD_LIBRARY_PATH 
	   environment variables.

	2. Install version 4.0 Release of the NVIDIA GPU Computing SDK by executing the file 
	   gpucomputingsdk_4.0_linux.run

	   The installer will prompt you to enter an installation path for the SDK or
	   accept the default.  We will refer to the path you choose as 
	   SDK_INSTALL_PATH.
	   
	3. Build the SDK project examples.  

		cd <SDK_INSTALL_PATH>/C
		make

	   Note Adding the following in make will build for specific targets
	 
			make x86_64=1     (for 64-bit targets)
			make i386=1       (for 32-bit targets)
	    
	4. Run the examples (32-bit or 64-bit Linux)
	    
		cd <SDK_INSTALL_PATH>/C/bin/linux/release
			matrixmul

	   (or any of the other executables in that directory)

	See the next section for more details on installing, building, and running
	SDK samples.

[Part 2/2]
	Note: The default installation folder <SDK_INSTALL_PATH> is "~/NVIDIA_GPU_Computing_SDK"

	This package consists of a ".run" file. This is a self-extracting archive that
	decompresses its contents to a temporary folder and then installs the contents 
	The archive is:

	gpucomputingsdk_4.0_linux.run: NVIDIA GPU Computing SDK Installer 

	In addition, a NVIDIA Linux Display driver is needed to run CUDA code on an 
	NVIDIA GPU.  CUDA 4.0 Release requires version 270 or newer version of the linux 
	NVIDIA Display Driver.  Please see the NVIDIA CUDA Toolkit 4.0 Release notes 
	for more details.

	   For information on installing NVIDIA Linux display drivers, please refer to 
	   the NVIDIA Accelerated Linux Driver Set README and Installation Guide:
	   http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9755/README/index.html

	1. Install version 4.0 Release of the NVIDIA CUDA Toolkit by executing the file
		 cudatoolkit_4.0_linux_*.run where * corresponds to your Linux distribution

	   To install, run the cudatoolkit_4.0_linux_*.run script.  You will be prompted
	   for the path to where you want to put the CUDA files. In the following we will 
	   call this path <CUDA_INSTALL_PATH>. It is recommended that you run the 
	   installer as root and use the default install path (/usr/local). 

	   Make sure that you add the location of the CUDA binaries (such as nvcc) to 
	   your PATH environment variable and the location of the CUDA libraries
	   (such as libcuda.so) to your LD_LIBRARY_PATH environment variable.

	   In the bash shell, one way to do this is to add the following lines to the 
	   file ~/.bash_profile from your home directory.

	   a. For 32-bit operating systems use the following paths 
		PATH=$PATH:<CUDA_INSTALL_PATH>/bin
		LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<CUDA_INSTALL_PATH>/lib
	    
	   b. For 64-bit operating systems use the following paths
		PATH=$PATH:<CUDA_INSTALL_PATH>/bin
		LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<CUDA_INSTALL_PATH>/lib64

	   Then to export the environment variables add this to the profile configuration
		export PATH
		export LD_LIBRARY_PATH
	 
	2. Install the NVIDIA GPU Computing SDK by executing the file gpucomputingsdk_4.0_linux.run

	   To install, run the gpucomputingsdk_4.0_linux.run script.  You will 
	   be prompted for the path to where you want to put the CUDA SDK.  You can 
	   regard the CUDA SDK as user code (it is a set of examples), and therefore
	   the default installation is in the current user's home directory 
	   (~/NVIDIA_GPU_Computing_SDK). You must either accept the default or specify a path
	   to which the user has write permissions. 

	   We will refer to the path you choose as SDK_INSTALL_PATH below.
	   
	3. Build the SDK project examples.  
		a. cd <SDK_INSTALL_PATH>/C
		b. Build:
			- release    configuration by typing "make".
			- debug      configuration by typing "make dbg=1".
			- x86_64=1   configuration by typing "make x86_64=1"
			- i386=1     configuration by typing "make i386=1"

		Running make at the top level first builds libcutil, a utility library used
		by the SDK examples (libcutil is simply for convenience -- it is not a part
		of CUDA and is not required for your own CUDA programs).  Make then builds
		each of the projects in the SDK.  

		NOTES:
		- The release and debug configurations require a CUDA-capable GPU to run
		  properly (see Appendix A.1 of the CUDA Programming Guide for a complete
		  list of CUDA-capable GPUs).

		- To build just libcutil, type "make" (or "make dbg=1") in the "common" 
		  subdirectory:

			cd <SDK_INSTALL_PATH>/C/common
			make

	4. Run the examples from the release or debug
	   directories located in 

			<SDK_INSTALL_PATH>/C/bin/linux/[release|debug].
and did you do this
Quote:

a. For 32-bit operating systems use the following paths
PATH=$PATH:<CUDA_INSTALL_PATH>/bin
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<CUDA_INSTALL_PATH>/lib

b. For 64-bit operating systems use the following paths
PATH=$PATH:<CUDA_INSTALL_PATH>/bin
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<CUDA_INSTALL_PATH>/lib64

Then to export the environment variables add this to the profile configuration
export PATH
export LD_LIBRARY_PATH

Last edited by John VV; 07-30-2011 at 02:26 AM.
 
Old 07-30-2011, 03:38 AM   #3
sag47
Senior Member
 
Registered: Sep 2009
Location: Orange County, CA
Distribution: Kubuntu x64, Raspbian, CentOS
Posts: 1,860
Blog Entries: 36

Rep: Reputation: 458Reputation: 458Reputation: 458Reputation: 458Reputation: 458
A friend of mine recently set up CUDA on his machine and he had to replace the video drivers with ones directly from nVidia. That could also potentially be the problem.
 
Old 07-30-2011, 05:13 AM   #4
John VV
LQ Muse
 
Registered: Aug 2005
Location: A2 area Mi.
Posts: 17,437

Rep: Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591
to add the sdk dose build with gcc 4.3
 
Old 07-31-2011, 08:29 PM   #5
bxsjc
LQ Newbie
 
Registered: Jul 2011
Posts: 5

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by John VV View Post
have you read the nvidia web site ? and the nvidia dev forum ?
http://developer.nvidia.com/category/zone/cuda-zone
from
http://developer.nvidia.com/cuda-toolkit-40#Linux
fedora 13 is the last supported for the cuda4 toolkit
it will require gcc 4.4 or 4.1
build it or it might be in the fedora repos
i do know that the old 3 built in gcc 4.1

right now i am DL'ing
h?p://developer.download.nvidia.com/compute/cuda/4_0/sdk/gpucomputingsdk_4.0.17_linux.run
BUT
i am running gcc4.5( or 4.3,4.1,or 3.4) and NOT NOT gcc 4.6
i do not think you can build the sdk using gcc4.6
4.6 is way WAY WAY to new .

did you read and fallow the "CUDA_SDK_Release_Notes.txt"
NVIDIA_GPU_Computing_SDK/doc/CUDA_SDK_Release_Notes.txt
Code:
--------------------------------------------------------------------------------
I. (b)  Linux Installation Instructions
--------------------------------------------------------------------------------
[Part 1/2]

	Note: The default installation folder <SDK_INSTALL_PATH> is "~/NVIDIA_GPU_Computing_SDK"

	For more detailed instructions, see section II below.

	0. Install the NVIDIA Linux display driver by executing the file

	   a. For 32-bit linux distributions use:
		  cudadriver_4.0_linux_32_270.xx.run

	   b. For 64-bit linux distributions use:
		  cudadriver_4.0_linux_64_270.xx.run
	  
	   For information on installing NVIDIA Linux display drivers, please refer to 
	   the NVIDIA Accelerated Linux Driver Set README and Installation Guide:
	   http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9755/README/index.html

	1. Install version 4.0 Release of the NVIDIA Toolkit by executing the file
	   cudatoolkit_4.0_linux_*.run where * corresponds to your Linux distribution

	   Add the CUDA binaries and lib path to your PATH and LD_LIBRARY_PATH 
	   environment variables.

	2. Install version 4.0 Release of the NVIDIA GPU Computing SDK by executing the file 
	   gpucomputingsdk_4.0_linux.run

	   The installer will prompt you to enter an installation path for the SDK or
	   accept the default.  We will refer to the path you choose as 
	   SDK_INSTALL_PATH.
	   
	3. Build the SDK project examples.  

		cd <SDK_INSTALL_PATH>/C
		make

	   Note Adding the following in make will build for specific targets
	 
			make x86_64=1     (for 64-bit targets)
			make i386=1       (for 32-bit targets)
	    
	4. Run the examples (32-bit or 64-bit Linux)
	    
		cd <SDK_INSTALL_PATH>/C/bin/linux/release
			matrixmul

	   (or any of the other executables in that directory)

	See the next section for more details on installing, building, and running
	SDK samples.

[Part 2/2]
	Note: The default installation folder <SDK_INSTALL_PATH> is "~/NVIDIA_GPU_Computing_SDK"

	This package consists of a ".run" file. This is a self-extracting archive that
	decompresses its contents to a temporary folder and then installs the contents 
	The archive is:

	gpucomputingsdk_4.0_linux.run: NVIDIA GPU Computing SDK Installer 

	In addition, a NVIDIA Linux Display driver is needed to run CUDA code on an 
	NVIDIA GPU.  CUDA 4.0 Release requires version 270 or newer version of the linux 
	NVIDIA Display Driver.  Please see the NVIDIA CUDA Toolkit 4.0 Release notes 
	for more details.

	   For information on installing NVIDIA Linux display drivers, please refer to 
	   the NVIDIA Accelerated Linux Driver Set README and Installation Guide:
	   http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9755/README/index.html

	1. Install version 4.0 Release of the NVIDIA CUDA Toolkit by executing the file
		 cudatoolkit_4.0_linux_*.run where * corresponds to your Linux distribution

	   To install, run the cudatoolkit_4.0_linux_*.run script.  You will be prompted
	   for the path to where you want to put the CUDA files. In the following we will 
	   call this path <CUDA_INSTALL_PATH>. It is recommended that you run the 
	   installer as root and use the default install path (/usr/local). 

	   Make sure that you add the location of the CUDA binaries (such as nvcc) to 
	   your PATH environment variable and the location of the CUDA libraries
	   (such as libcuda.so) to your LD_LIBRARY_PATH environment variable.

	   In the bash shell, one way to do this is to add the following lines to the 
	   file ~/.bash_profile from your home directory.

	   a. For 32-bit operating systems use the following paths 
		PATH=$PATH:<CUDA_INSTALL_PATH>/bin
		LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<CUDA_INSTALL_PATH>/lib
	    
	   b. For 64-bit operating systems use the following paths
		PATH=$PATH:<CUDA_INSTALL_PATH>/bin
		LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<CUDA_INSTALL_PATH>/lib64

	   Then to export the environment variables add this to the profile configuration
		export PATH
		export LD_LIBRARY_PATH
	 
	2. Install the NVIDIA GPU Computing SDK by executing the file gpucomputingsdk_4.0_linux.run

	   To install, run the gpucomputingsdk_4.0_linux.run script.  You will 
	   be prompted for the path to where you want to put the CUDA SDK.  You can 
	   regard the CUDA SDK as user code (it is a set of examples), and therefore
	   the default installation is in the current user's home directory 
	   (~/NVIDIA_GPU_Computing_SDK). You must either accept the default or specify a path
	   to which the user has write permissions. 

	   We will refer to the path you choose as SDK_INSTALL_PATH below.
	   
	3. Build the SDK project examples.  
		a. cd <SDK_INSTALL_PATH>/C
		b. Build:
			- release    configuration by typing "make".
			- debug      configuration by typing "make dbg=1".
			- x86_64=1   configuration by typing "make x86_64=1"
			- i386=1     configuration by typing "make i386=1"

		Running make at the top level first builds libcutil, a utility library used
		by the SDK examples (libcutil is simply for convenience -- it is not a part
		of CUDA and is not required for your own CUDA programs).  Make then builds
		each of the projects in the SDK.  

		NOTES:
		- The release and debug configurations require a CUDA-capable GPU to run
		  properly (see Appendix A.1 of the CUDA Programming Guide for a complete
		  list of CUDA-capable GPUs).

		- To build just libcutil, type "make" (or "make dbg=1") in the "common" 
		  subdirectory:

			cd <SDK_INSTALL_PATH>/C/common
			make

	4. Run the examples from the release or debug
	   directories located in 

			<SDK_INSTALL_PATH>/C/bin/linux/[release|debug].
and did you do this



If you know any errors in the following, please tell me. Thank you!
Process of Installing CUDA and CUDA Toolkit
(I am just a newbie and want to share my experience on CUDA installation. So if you know any errors in my installation, please tell me : bxsjc0728@126.com or bxsjc0728@hotmail.com .Thank you!)
As a newbie, I have spent 10 days to install CUDA and the rest in connection with it. No pains, no gains!
Now, let me put down the correct process of installation. I wish everyone could make it quickly.
Download fedora 15 from http://fedoraproject.org/ . I use this one : http://download.fedoraproject.org/pu...6-Live-KDE.iso .
When the installation of fedora 15 is done, I run the following commands:
yum install firefox.i686(version—5.0-2.fc15)
yum install stardict.i686(version—3.0.2-2.fc15)
yum install fcitx.i686(version—4.0.1-3.fc15)
Then I went to http://rpmfusion.org/ and downloaded two rpms:
http://download1.rpmfusion.org/free/...ble.noarch.rpm http://download1.rpmfusion.org/nonfr...ble.noarch.rpm
OK, the next step is to run these commands:
yum install kernel.i686(version: 2.6.38.8-35.fc15)
yum install kernel-devel.i686(version: 2.6.38.8-35.fc15)
yum install binutils.i686(version: 2.21.51.0.6-6.fc15)
yum install gcc.i686(version: 4.6.0-9.fc15)
NOTICE: you must match kernel-devel.i686 and kernel.i686.
Now, download NVIDIA driver suitable for your system by entering http://www.nvidia.com/Download/index.aspx?lang=en-us .
Then, download cudatoolkit 4.0 and gpucomputingsdk 4.0 by turning to http://developer.nvidia.com/cuda-toolkit-40 . Please select the right one and download.
OK, now I have NVIDIA-Linux-x86-275.21.run ; gpucomputingsdk_4.0.17_linux.run ; cudatoolkit_4.0.17_linux_32_fedora13.run in /home/bxsjc/Downloads/ .
NOTICE: you might have a doubt about fedora13.run .you know my system is fedora15. I think that's not a problem for CUDA. In fact, I haven't received any error report related to this difference. If you know anything about this, please send e-mail to bxsjc0728@126.com or bxsjc0728@hotmail.com .Thank you!
Let's run this command “ln -sf /lib/systemd/system/runlevel3.target /etc/systemd/system/default.target”.
Then, reboot the system, we will enter the command line instead of X window. That's just what NVIDIA driver wants! If you do not such command ,you will know you are wrong and I am right. It's up to you.
(For this step, you can also “ Ctrl+Alt+F1/F2/.../F6 ”, login as root and do “ init 3 ”.)
OK, now we are in the command line and we log in as root. However , do nor hurry to install NVIDIA driver! We still have something to do. Run these commands:
vi /etc/modprobe.d/blacklist.conf
/# add “ blacklist nouveau ” and add “#” before “ blacklist nvidiafb ”
vi /boot/grub/grub.conf
/# add “ rbblacklist=nouveau ” behind the first “ rhgb quiet ”
sudo mv /boot/initramfs-$(uname -r).img /boot/initramfs-$(uname -r).img.bak
sudo dracut -v /boot/initramfs-$(uname -r).img $(uname -r)
OK, now we make the default driver nouveau not working which is also what NVIDIA driver requires!
Reboot . Move NVIDIA-Linux-x86-275.21.run ; gpucomputingsdk_4.0.17_linux.run ; cudatoolkit_4.0.17_linux_32_fedora13.run to ~/ .Then code:
cd ~
./ NVIDIA-Linux-x86-275.21.run
/#everything will be ok.
ln -sf /lib/systemd/system/runlevel5.target /etc/systemd/system/default.target
OK, reboot and we will return to X window. Of course, X window is more friendly to us.
(For this step, you can also do “ init 5 ” as root.)
However, we still run the terminal and “ yum install compat-gcc-34.i686(version: 3.4.6-22.fc15) ”
and “ yum install compat-gcc-34-c++.i686(version: 3.4.6-22.fc15) ”. Why do I install compat-gcc-34? Because cudatoolkit 4.0 do not support gcc4.6 . However, installing these two packages is enough? Of course NO! You need to input these codes:
mv /usr/bin/gcc /usr/bin/gcc46
ln -sf /usr/bin/gcc34 /usr/bin/gcc
ln -sf /usr/bin/g++34 /usr/bin/g++
These efforts can set compat-gcc-34 as the dafault gcc version.(Looking for the right code took me much time!)
Now, “ cd ~ ” and “ ./cudatoolkit_4.0.17_linux_32_fedora13.run ”. When the installation finishes, there will be some error reports, like these sentences:
Please make sure your PATH includes /usr/local/cuda/bin
Please make sure your LD_LIBRARY_PATH
for 32-bit Linux distributions includes /usr/local/cuda/lib
for 64-bit Linux distributions includes /usr/local/cuda/lib64:/usr/local/cuda/lib
or
for 32-bit Linux distributions add /usr/local/cuda/lib
for 64-bit Linux distributions add /usr/local/cuda/lib64 and /usr/local/cuda/lib
to /etc/ld.so.conf and run ldconfig as root.
So, act as told! “ vi ~/.bash_profile ” as root and bxsjc(my name)and the content needs to be as followings:
“ PATH=$PATH:$HOME/bin:/usr/local/cuda/bin
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib
export PATH
export LD_LIBRARY_PATH
”
(Or “ vi /etc/ld.so.conf ” as root, you need to add “ /usr/local/cuda/lib ”. Then , “ sudo ldconfig ”)
(Maybe I should do this instead of the above one:“ cat > /etc/ld.so.conf.d/cuda.conf ”, the content needs to be “ /usr/local/cuda/lib ”)
When these steps finish, install your gpucomputingsdk.run .Maybe there is still some reports saying that please make sure something .Ignore them! You are not its slaves .
I am sorry to tell you that there are still some shits to deal with. Please be patient. Thank you!
“
yum install libtool.i686(version: 2.4-4.fc15)
yum install freeglut.i686(version: 2.6.0-6.fc15)
yum install libX11-devel.i686(version: 1.4.3-1.fc15)
ln -s /usr/lib/libglut.so.3 /usr/lib/libglut.so
ln -s /usr/lib/libGLU.so.1 /usr/lib/libGLU.so
ln -s /usr/lib/libX11.so.6 /usr/lib/libX11.so
ln -s /usr/lib/libXi.so.6 /usr/lib/libXi.so
ln -s /usr/lib/libXmu.so.6 /usr/lib/libXmu.so
”
Congratulations! Now you can “ cd ~/NVIDIA_GPU_Computing_SDK/C/ ” and “ make ”.
Enjoy it! There will be no errors!
Maybe there some warnings , but according to my results, these warnings do not affect files in ~/NVIDIA_GPU_Computing_SDK/C/bin/linux/release . If you know anything about this problem, please tell me(e-mail: bxsjc0728@126.com or bxsjc0728@hotmail.com ). Thank you!
 
Old 07-31-2011, 09:03 PM   #6
John VV
LQ Muse
 
Registered: Aug 2005
Location: A2 area Mi.
Posts: 17,437

Rep: Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591
Quote:
As a newbie, I have spent 10 days to install CUDA and the rest in connection with it. No pains, no gains!
cuda DOSE NOT build the examples using fedora 15 's gcc 4.6 or 4.5
you MUST use gcc 4.3 or 4.1

-- fact of life with fedora --
-- it is VERY often way WAY WAY to new for most software --
-- fedora is a Research and Development /testing distro --
-- it is NOT new to linux friendly ( often it is using WAY too new of code for most software ) --
-- live with it or use a different distro --

Quote:
yum install firefox.i686(version—5.0-2.fc15)
why
firefox is already installed by default
there is no need to reinstall it .
Quote:
yum install kernel-devel.i686(version: 2.6.38.8-35.fc15)
yum install binutils.i686(version: 2.21.51.0.6-6.fc15)
yum install gcc.i686(version: 4.6.0-9.fc15)
why
the recommended way with all red hat types is a group install
Code:
su -
yum groupinstall " Development Tools" "Development Libraries"
with "binutils" and gcc you are missing things

Quote:
yum install gcc.i686(version: 4.6.0-9.fc15)
that version will NOT work for cuda -- it is way too new
use gcc 4.3
and DO NOT use the very old legacy 3.4

Quote:
However, we still run the terminal and “ yum install compat-gcc-34.i686(version: 3.4.6-22.fc15) ”
do not call that one and one only program by it's name
Code:
su -
yum grouplist
then from that list install the "legacy development tools "
there are a whole mess of things that 3.4 must have installed
--- do not build cuda with 3.4

instal gcc 4.3
-- yes that is a pain on fedora
fedora's gcc is way too new
you might have to build 4.3 from source and install it side by side with 4.5
-- google that , the instructions are way too long for this thread


Quote:

yum install libtool.i686(version: 2.4-4.fc15)
yum install freeglut.i686(version: 2.6.0-6.fc15)
yum install libX11-devel.i686(version: 1.4.3-1.fc15)
why ?
libtool IS installed from the group install that you did not do



---------- you might have to start over ??? maybe ? may be not -------------


please use the groupinstall's that i posted above
for " Development Tools" "Development Libraries"
build gcc4.3 FROM SOURCE
or use a different distro fedora's gcc 4.5 is TOO NEW for cuda
but it might be in fedora's repos
Code:
su -
yum search gcc
if there that will show it
 
Old 07-31-2011, 09:13 PM   #7
bxsjc
LQ Newbie
 
Registered: Jul 2011
Posts: 5

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by John VV View Post
-- fact of life with fedora --
-- it is VERY often way WAY WAY to new for most software --
-- fedora is a Research and Development /testing distro --
-- it is NOT new to linux friendly ( often it is using WAY too new of code for most software ) --
-- live with it or use a different distro --
Thank you! You are my teacher for life! Thank you! I will do it again and better!
 
Old 08-01-2011, 10:38 PM   #8
John VV
LQ Muse
 
Registered: Aug 2005
Location: A2 area Mi.
Posts: 17,437

Rep: Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591Reputation: 2591
cross posted on fedoraforum
http://forums.fedoraforum.org/showthread.php?t=267629
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
Installing CUDA SDK on Fedora stuart_m LinuxQuestions.org Member Success Stories 0 02-21-2011 05:43 AM
Issue with CUDA-BLASTP abhicdac Linux - Software 2 12-22-2010 02:09 PM
Issue with CUDA-BLASTP abhicdac Linux - General 2 12-22-2010 03:56 AM
[SOLVED] cuda for gfortran ? aihaike Programming 4 09-26-2010 08:07 PM
CUDA on 11.2 (64 bit) marios_auth SUSE / openSUSE 1 04-02-2010 10:46 PM

LinuxQuestions.org > Forums > Non-*NIX Forums > Programming

All times are GMT -5. The time now is 04:47 AM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration