Thursday, March 08, 2012

HOWTO : BackTrack 5 R2 on Intel X79 Express and nVidia display cards chipset

UPDATE : The better way is here.

Hardware

CPU : Intel i7-3930K (Socket 2011, 12 cores with HT)
Display card : 2 x nVidia GeForce GTX 590 (1024 CUDA cores per card)

Installation of BackTrack 5 R2

BackTrack 5 R2 can be boot up on Intel X79 Express chipset motherboard with 2 nVidia GeForce GTX 590 display cards. However, "nomodeset" should be applied to the boot option by pressing "tab" on the boot menu.

Install the BackTrack 5 R2 as usual. When it is required to reboot, do not remove the BackTrack 5 R2 CD. Boot up the CD accordingly. After the BackTrack 5 R2 is booted up, mount the hard drive and add "nomodeset" to boot option of the grub.cfg at /boot/grub.

After that, reboot the system and remove the CD. The system will be boot into BackTrack 5 R2 without problem.

If the kernel is upgraded, you should rebuild the kernel headers by the following steps :

prepare-kernel-sources
cd /usr/src/linux
cp -rf include/generated/* include/linux/


Installation of nVidia display driver

Go to nVidia Deleloper Zone CUDA Toolkit 4.1 to download the following. Do not enter to X11 by issuing "startx"; otherwise, the installation will be failed.

(1) Download "Latest Drivers for Linux (295.20)" for the nVidia Driver. You can download the driver before not going to the X11.

32-bit :
http://www.nvidia.com/object/linux-display-ia32-295.20-driver.html

64-bit :
http://www.nvidia.com/object/linux-display-amd64-295.20-driver.html

Installation :

32-bit :
chmod +x NVIDIA-Linux-x86-295.20.run
./NVIDIA-Linux-x86-295.20.run


64-bit :
chmod +x NVIDIA-Linux-x86_64-295.20.run
./NVIDIA-Linux-x86_64-295.20.run


After that, you can reboot your system.

Then run the following command to make the suitable resolution :
nvidia-xconfig

(2) Download "CUDA Toolkit for Ubuntu Linux 10.04" for the CUDA Toolkit.

32-bit :
wget http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/cudatoolkit_4.1.28_linux_32_ubuntu10.04.run

64-bit :
wget http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/cudatoolkit_4.1.28_linux_64_ubuntu10.04.run

chmod +x http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/cudatoolkit_4.1.28_linux_xx_ubuntu10.04.run

*** Accept the default settings.

(3) Download "GPU Computing SDK" for the nVidia SDK.

wget http://developer.download.nvidia.com/compute/cuda/4_1/rel/sdk/gpucomputingsdk_4.1.28_linux.run

chmod +x gpucomputingsdk_4.1.28_linux.run
./gpucomputingsdk_4.1.28_linux.run


nano /root/.bashrc

Append the following :

PATH=$PATH:/usr/local/cuda/bin
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib:/usr/local/cuda/lib64
export PATH
export LD_LIBRARY


source /root/.bashrc
ldconfig


After that, reboot the system to make the nVidia driver effect.

Compile Sample code of CUDA

apt-get install freeglut3-dev libxi-dev libXmu-dev

cd NVIDIA_GPU_computing_SDK/C
make


Then, run the sample codes at :

cd NVIDIA_GPU_computing_SDK/C/bin/linux/release
./deviceQuery
./nbody


Installation of pyrit

Add the following to /etc/apt/sources.list :

deb http://security.ubuntu.com/ubuntu lucid-security main
apt-get update
apt-get libssl-dev scapy python-dev

Then, delete the newly added entry of the sources.list.

apt-get update

Go to the official site of pyrit.

cd ~
svn checkout http://pyrit.googlecode.com/svn/trunk/ pyrit

cd pyrit
python setup.py build
python setup.py install


cd cpyrit
python setup.py build
python setup.py install


To test if the installation is correct or not.

pyrit list_cores
pyrit benchmark
pyrit benchmark_long


Remarks :

Fix the splash :
fix-splash

Install the sensors :
apt-get install sensors-applet

Pyrit Cluster Setup
John the Ripper Cluster Setup

See Also : HOWTO : Pyrit Cluster with BackTrack 5 R2

*** Whenever the linux kernel is updated or upgraded, the nVidia driver should be required to reinstalled to match the upgraded kernel version.

That's all! See you.

Before water cooling



After water cooling