search

JPaulMora / Pyrit

The famous WPA precomputed cracker, Migrated from Google.
GNU General Public License v3.0
1.07k stars 273 forks source link

How to Install Pyrit Cuda + Toolkit 8 + Nvidia GTX 1070 in Kali Linux 2017.1 #528

Open reproteq opened 7 years ago

reproteq commented 7 years ago

Install Kali Linux 2017.1 + Nvidia GTX-1070 drivers + Toolkit 8 + Pyrit Cuda for computing Gpu Cores

Descargar y dar permisos

https://developer.nvidia.com/cuda-downloads 

chmod +x cuda_8.0.61_375.26_linux.run

http://www.nvidia.com/download/index.aspx?lang=es 

chmod +x NVIDIA-Linux-x86_64-375.66.run

Editar fuentes

leafpad /etc/apt/sources.list                                                              
deb http://http.kali.org/kali kali-rolling main non-free contrib    
deb-src http://http.kali.org/kali kali-rolling main non-free contrib

Actualizar distro e instalar headers

apt-get update                                                                                 
apt-get dist-upgrade    
apt-get install -y linux-headers-$(uname -r)

Desactivar driver nouveau y actualizar grub

sed 's/quiet/quiet nouveau.modeset=0/g' -i /etc/default/grub    
update-grub 
reboot

Modo consola sin entorno grafico

ctrl+alt+f3
login como root

Instalar driver nvidia

./NVIDIA-Linux-x86_64-375.66.run                                                 
update-grub 
reboot

Instalar cuda toolkit

export PERL5LIB=.                                                                     
./cuda_8.0.61_375.26_linux.run --override

Muestra paths

ldconfig -v 2>/dev/null | grep -v ^$'\t'

Editar config paths

leafpad /etc/ld.so.conf
## incluir esta linea en ld.so.conf                                                     
/usr/local/cuda-8.0/lib64

Instala librerias

apt-get install python2.7-dev libssl-dev zlib1g-dev libpcap-de

Clonar repositorio

git clone https://github.com/JPaulMora/Pyrit.git

Construir instalador e instalar

python setup.py clean build install

Activar CUDA en config

leafpad /root/.pyrit/config                                                                     
use_CUDA = true 
pyrit list_cores                                                                                       

pyrit benchmark

Computed 118727.81 PMKs/s total.
#1: 'CPU-Core (SSE2/AES)': 570.0 PMKs/s (RTT 2.9)
#2: 'CPU-Core (SSE2/AES)': 563.3 PMKs/s (RTT 3.0)
#3: 'CPU-Core (SSE2/AES)': 575.7 PMKs/s (RTT 3.0)
#4: 'CPU-Core (SSE2/AES)': 578.7 PMKs/s (RTT 3.0)
#5: 'CPU-Core (SSE2/AES)': 567.0 PMKs/s (RTT 3.0)
#6: 'CPU-Core (SSE2/AES)': 535.9 PMKs/s (RTT 2.9)
#7: 'CPU-Core (SSE2/AES)': 551.7 PMKs/s (RTT 2.9)
#8: 'CPU-Core (SSE2/AES)': 560.9 PMKs/s (RTT 3.1)
CUDA:
#1: 'CUDA-Device #1 'GeForce GTX 1070'': 127866.4 PMKs/s (RTT 1.0)
JPaulMora commented 7 years ago

Thanks, I'll edit nvcc_cmd = NVCC + bit_flag + ' -ccbin clang' ' ./_cpyrit_cudakernel.cu' directly in the source

reproteq commented 7 years ago

Thanks and sorry this is the correct code else errors.

        nvcc_cmd = NVCC + bit_flag + ' -ccbin clang'\
                                     ' -Xcompiler "-fPIC" --ptx' \
                                     ' ./_cpyrit_cudakernel.cu'
JPaulMora commented 7 years ago

ok

emrysrind commented 7 years ago

all works well but I cannot install toolkit its been a while that I am trying to install it at every end it shows same missing library message which wasnt there before 8.0 version sadly not even cuda 7.5 is working anymore gives the same error any help/suggestions ?

Description

This package includes over 100+ CUDA examples that demonstrate various CUDA programming principles, and efficient CUDA implementation of algorithms in specific application domains. The NVIDIA CUDA Samples License Agreement is available in Chapter 2. Do you accept the previously read EULA? accept/decline/quit: accept

You are attempting to install on an unsupported configuration. Do you wish to continue? (y)es/(n)o [ default is no ]: y

Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 375.26? (y)es/(n)o/(q)uit: n

Install the CUDA 8.0 Toolkit? (y)es/(n)o/(q)uit: y

Enter Toolkit Location [ default is /usr/local/cuda-8.0 ]:

Do you want to install a symbolic link at /usr/local/cuda? (y)es/(n)o/(q)uit: y

Install the CUDA 8.0 Samples? (y)es/(n)o/(q)uit: y

Enter CUDA Samples Location [ default is /root ]:

Installing the CUDA Toolkit in /usr/local/cuda-8.0 ... Missing recommended library: libXi.so Missing recommended library: libXmu.so

Warning: cannot find Toolkit in /usr/local/cuda-8.0. Use --toolkitpath to specify the toolkit location. Installing the CUDA Samples in /root ... sh: 1: /usr/local/cuda-8.0/bin/cuda-install-samples-8.0.sh: not found chown: cannot access '/root/NVIDIA_CUDA-8.0_Samples': No such file or directory

=========== = Summary =

Driver: Not Selected Toolkit: Installation Failed Samples: Installation Failed

Logfile is /tmp/cuda_install_2039.log

JPaulMora commented 7 years ago

Weird, this seems like a bug with the toolkit installer.. why don't you try installing from the default repositories for your distro?

emrysrind commented 7 years ago

you mean apt-get install nvidia-cuda-toolkit nvidia-opencl-icd

JPaulMora commented 7 years ago

yes

reproteq commented 7 years ago

download cuda 8 from nvidia develorpers type: export PERL5LIB=.
and install cudatoolkit type:
./cuda_8.0.61_375.26_linux.run --override
and acept everythings but not driver because driver nvidia i install previusly

reproteq commented 7 years ago

pyrit cudas is working in kali linux 2017.1 and last driver nvidia and toolkit 8.xx i tested and work 100% friend gpu compute aprox 130.000 pmks

h3xcmd commented 7 years ago

Success with the nvidia-cuda-toolkit install from repros on a Dell Latitude Optimus (intel+nvidia) laptop. Clean install of Kali 2017.1 and stay logged into ROOT

Dell Latitude 14 Rugged Extreme (7404) Fourth-generation Intel Core™ i7-4650U @ 4x 3.3GHz GPU #1: Integrated Intel HD Graphics 5000 GPU #2: NVIDIA GeForce GT 720M (DDR3 2GB) discrete graphics with Optimus™

Guides that helped me

Clean Install Kali 2017.3 Kernel 4.13.0-kali1-amd64 and Update

apt update apt upgrade apt autoremove apt -f install apt upgrade

Remove the installed Kali Pyrit

apt remove pyrit

Nvidia Driver prep

Verify nvidia

lspci | grep -E "VGA|3D"

Blacklist Nouveau

echo -e "blacklist nouveau\noptions nouveau modeset=0\nalias nouveau off" > /etc/modprobe.d/blacklist-nouveau.conf

update-initramfs -u && reboot

Verify nouveau is disabled

lsmod | grep -i nouveau

must be a blank return

[2GB install] Automatic Install

apt install nvidia-driver ocl-icd-libopencl1 nvidia-cuda-toolkit

Additional install

apt install mesa-utils nvidia-cuda-gdb nvidia-cuda-doc libssl-dev zlib1g-dev libpcap-dev libgl1-nvidia-glx python2.7-dev python-pycuda-doc python-pycuda python3-pycuda

NEXT

nvidia-xconfig --query-gpu-info | grep 'BusID : ' | cut -d ' ' -f6 PCI:3:0:0

Replace the bold string with your Bus ID and save it to /etc/X11/xorg.conf

gedit /etc/X11/xorg.conf

Section "ServerLayout" Identifier "layout" Screen 0 "nvidia" Inactive "intel" EndSection

Section "Device" Identifier "nvidia" Driver "nvidia" BusID "PCI:3:0:0" EndSection

Section "Screen" Identifier "nvidia" Device "nvidia" Option "AllowEmptyInitialConfiguration" EndSection

Section "Device" Identifier "intel" Driver "modesetting" EndSection

Section "Screen" Identifier "intel" Device "intel" EndSection

Create these two files and add the same script to each

gedit /usr/share/gdm/greeter/autostart/optimus.desktop gedit /etc/xdg/autostart/optimus.desktop

[Desktop Entry]
Type=Application
Name=Optimus
Exec=sh -c "xrandr --setprovideroutputsource modesetting NVIDIA-0; xrandr --auto"
NoDisplay=true
X-GNOME-Autostart-Phase=DisplayServer

Reboot, Pray, Verify

glxinfo | grep -i "direct rendering" yes

Fix screen tearing (optional)

xrandr --verbose|grep PRIME gedit /etc/default/grub

GRUB_CMDLINE_LINUX_DEFAULT="quiet nvidia-drm.modeset=1"

update-grub reboot

xrandr --verbose|grep PRIME All should be 1's <-this just prevents screen tearing

Pyrit prep

Added these lines to the end of .bashrc

gedit /root/.bashrc

PATH=$PATH:/usr/lib/nvidia-cuda-toolkit/bin.
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/nvidia-cuda-toolkit/lib:/lib
export PATH
export LD_LIBRARY_PATH

Pyrit Install

Extract contents of Pyrit-master.zip to /OPT directory (its just easier)

cd /opt/Pyrit-master/
python setup.py clean build install

Pyrit_Cuda Prep and Install

cd /opt/Pyrit-master/modules/cpyrit_cuda/ gedit setup.py

Edit the setup.py, line 35**,edit it from

for path in ('/usr/local/cuda', '/opt/cuda'):

to

for path in ('/usr/local/cuda','/usr/lib/nvidia-cuda-toolkit','/opt/cuda'):

So the nvcc could be found during the process,then install

Possible Errors

Their could be a possible error when you run the /opt/Pyrit-master/modules/cpyrit_cuda/python setup.py install if you get the bellow error: 

usr/bin/../lib/gcc/x86_64-linux-gnu/7.2.0/../../../../include/c++/7.2.0/bits/std_abs.h(101): error: identifier "__float128" is undefined
/usr/bin/../lib/gcc/x86_64-linux-gnu/7.2.0/../../../../include/c++/7.2.0/bits/std_abs.h(102): error: identifier "__float128" is undefined
2 errors detected in the compilation of "/tmp/tmpxft_00005498_00000000-7__cpyrit_cudakernel.cpp1.ii".

in the file /usr/include/c++/7/bits/std_abs.h line 99-104 . 

#if !defined(__STRICT_ANSI__) && defined(_GLIBCXX_USE_FLOAT128)
  inline _GLIBCXX_CONSTEXPR
  __float128
  abs(__float128 __x)
  { return __x < 0 ? -__x : __x; }
#endif

I Patched it on line 99 removing the current line and adding the string below in its place

#if defined(_GLIBCXX_USE_FLOAT128) && !defined(__STRICT_ANSI__) && !defined(__CUDACC__)
  inline _GLIBCXX_CONSTEXPR

Another Error I have personally experienced with the latest cuda toolkit fir the install for /cpyrit_cuda# python setup.py clean build install

Compiling CUDA module using nvcc 8.0, V8.0.61...
Executing '/usr/local/cuda/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc fatal : The version ('40001') of the host compiler ('clang') is not supported
Traceback (most recent call last):
File "setup.py", line 170, in 
setup(**setup_args)
File "/usr/lib/python2.7/distutils/core.py", line 151, in setup
dist.run_commands()
File "/usr/lib/python2.7/distutils/dist.py", line 953, in run_commands
self.run_command(cmd)
File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command
cmd_obj.run()
File "/usr/lib/python2.7/distutils/command/build.py", line 128, in run
self.run_command(cmd_name)
File "/usr/lib/python2.7/distutils/cmd.py", line 326, in run_command
self.distribution.run_command(command)
File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command
cmd_obj.run()
File "setup.py", line 93, in run
subprocess.check_call(nvcc_cmd, shell=True)
File "/usr/lib/python2.7/subprocess.py", line 190, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '/usr/local/cuda/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' returned non-zero exit status 1

The Fix is to

cd /opt/Pyrit/modules/cpyrit_cuda/ gedit setup.py

line 88,edit it from**

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

to

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang-3.8 -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

DO NOT FORGET TO:

gedit /root/.pyrit/config

default_storage = file:// limit_ncpus = 0 rpc_announce = true rpc_announce_broadcast = false rpc_knownclients = rpc_server = false use_CUDA = true use_OpenCL = false workunit_size = 75000

RESULTS

root@Kali:/opt/Pyrit-master/modules/cpyrit_cuda\% pyrit benchmark Pyrit 0.5.1 (C) 2008-2011 Lukas Lueg - 2015 John Mora https://github.com/JPaulMora/Pyrit This code is distributed under the GNU General Public License v3+

Running benchmark (6641.4 PMKs/s)... |

Computed 6641.36 PMKs/s total.

1: 'CPU-Core (SSE2/AES)': 295.9 PMKs/s (RTT 3.3)

2: 'CPU-Core (SSE2/AES)': 300.3 PMKs/s (RTT 3.1)

3: 'CPU-Core (SSE2/AES)': 272.6 PMKs/s (RTT 3.2)

4: 'CPU-Core (SSE2/AES)': 276.7 PMKs/s (RTT 3.2)

CUDA:

1: 'CUDA-Device #1 'GeForce GT 720M'': 6171.1 PMKs/s (RTT 2.8)

HaxorGaruda commented 7 years ago

nice tutorial @h3xcmd ... thanks so much ... i just follow edit the std_abs.h and setup.py on cpyrit_cuda, everythings work fine ... but i don't follow to install legacy driver ... because official distro has different way : https://www.kali.org/news/cloud-cracking-with-cuda-gpu/ and if i follow your step, it will be remove arch 32 on my kali linux 64 bit ... (ex : wine32) thanks...

h3xcmd commented 7 years ago

Your Welcome @HaxorGaruda As much as I struggled with this for a whole year on my Optimus Dell Latitude (intel+nvidia) I really wish it was as simple as

apt-get install -y ocl-icd-libopencl1 nvidia-driver nvidia-cuda-toolkit

I too thought it was a bit overkill to include nvidia legacy and the long list of installs for the nvidia driver step and you are right about something, breaking the system if you install or have 32bit arch is eminent with nvidia 64. After enabling 386 arch dpkg --add-architecture i386 && apt-get update it runs through the rolling repositories it tries to install 32bit nvidia drivers and that just messes up the mods with xorg.conf stopping gdm3 from loading. Through my trial and error with an Optimus system I have learned to just not use 32bit arch.

If you have discrete Nvidia hardware, I do not see a problem using the simplified install from the Kali Docs and yes your method would work. But in an Optimus laptop something is missing where it needs to switch from intel to nvidia and to be able to disable that switching ability to benchmark pyrit_cuda is what I came up with.

After figuring out what FLOAT_128 meant, the std_abs.h file was the most important step out of all of this besides figuring out the cocktail of installs to do.

pwnhun73r commented 7 years ago

hi, i have a problem when run "pyrit list_cores" only show me the processor cores. i have a gtx 1060 video card. are there other steps for install? thanks

Taslack commented 7 years ago

@c3ph3u5qwerty If you are on a laptop you need to use optirun pyrit list_cores, ect. ect.

pwnhun73r commented 7 years ago

@Taslack i have a desktop pc, i follow step by step but pyrit dont recognise the gpu core

h3xcmd commented 7 years ago

@c3ph3u5qwerty I am going to edit my answer even though its geared to Optimus Laptops that want to ditch the Intel display driver and go 100% with nvidia without auto switching with bumblebee where you do not need to run optirun. Also you have to run this tool as root and did you forget to edit the /root/.pyrit/config file cuda=true

@Taslack Optirun is for bumblebee-nvidia-driver Optimus laptops where you would want to use your default Intel display drivers and when needed to auto switch to Nvidia to accelerate your graphics. and in the past I never had any luck whatsoever with bumblebee and pyrit. You absolutely need the nvidia-drivers.

On this last fresh install with 2017.2 updated to Kali 2017.3; installing Pyrit, I did not need to edit the FLOAT_128 integer and I did not have to install a block of librarys and nvidia stuff previously mentioned. Its all included in the apt install nvidia-driver ocl-icd-libopencl1 nvidia-cuda-toolkit

Andiansyah23 commented 6 years ago

HI I NEED HELP!! ASAP

root@0mount:/0mount/wifi_attack/Pyrit/modules/cpyrit_cuda# python setup.py clean build install running clean Removing temporary files and pre-built GPU-kernels... running build running build_ext Compiling CUDA module using nvcc 8.0, V8.0.61... Executing '/usr/local/cuda/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning). nvcc fatal : The version ('40001') of the host compiler ('clang') is not supported Traceback (most recent call last): File "setup.py", line 170, in setup(**setup_args) File "/usr/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/usr/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/usr/lib/python2.7/distutils/command/build.py", line 128, in run self.run_command(cmd_name) File "/usr/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "setup.py", line 93, in run subprocess.check_call(nvcc_cmd, shell=True) File "/usr/lib/python2.7/subprocess.py", line 190, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '/usr/local/cuda/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' returned non-zero exit status 1

kudzie567 commented 6 years ago

Also encountering the error below after python setup.py clean build install when installing cpyrit_cuda. Here is the error: python /opt/Pyrit-master/modules/cpyrit_cuda/setup.py clean build install running build running build_ext Compiling CUDA module using nvcc 8.0, V8.0.61... Executing '/usr/lib/nvidia-cuda-toolkit/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning). nvcc fatal : The version ('40001') of the host compiler ('clang') is not supported Traceback (most recent call last): File "/opt/Pyrit-master/modules/cpyrit_cuda/setup.py", line 166, in setup(**setup_args) File "/usr/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/usr/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/usr/lib/python2.7/distutils/command/build.py", line 128, in run self.run_command(cmd_name) File "/usr/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/opt/Pyrit-master/modules/cpyrit_cuda/setup.py", line 91, in run subprocess.check_call(nvcc_cmd, shell=True) File "/usr/lib/python2.7/subprocess.py", line 190, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '/usr/lib/nvidia-cuda-toolkit/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' returned non-zero exit status 1

Hackitfree commented 6 years ago

Finally made it work on RYZEN 5 1600 with NVIDIA GTX 1060 3G...Getting 125 000 PMK/s (1CPU core+GPU).. I was getting errors like all of you but since i changed ( clang to clang-3.8) in setup.py in root@kali:/Pyrit/modules/cpyrit_cuda# everything works well.. Thank you guys for this thread !!!! Note: First you uninstall (apt remove pyrit) from Kali Then compile setup.py in root@kali:/Pyrit/ then compile again setup.py in root@kali:/Pyrit/modules/cpyrit_cuda# ...Hope i can help with this :)

PastiiJay commented 6 years ago

Thank you all! Nice work!

root@KaliDesktop:/opt/Pyrit-master/modules/cpyrit_cuda# pyrit benchmark Pyrit 0.5.1 (C) 2008-2011 Lukas Lueg - 2015 John Mora https://github.com/JPaulMora/Pyrit This code is distributed under the GNU General Public License v3+

Running benchmark (149966.0 PMKs/s)... /

Computed 149965.98 PMKs/s total.

1: 'CPU-Core (SSE2/AES)': 600.7 PMKs/s (RTT 3.0)

2: 'CPU-Core (SSE2/AES)': 608.6 PMKs/s (RTT 3.0)

3: 'CPU-Core (SSE2/AES)': 602.4 PMKs/s (RTT 2.9)

4: 'CPU-Core (SSE2/AES)': 625.4 PMKs/s (RTT 2.9)

5: 'CPU-Core (SSE2/AES)': 591.3 PMKs/s (RTT 3.0)

6: 'CPU-Core (SSE2/AES)': 604.2 PMKs/s (RTT 2.9)

7: 'CPU-Core (SSE2/AES)': 614.4 PMKs/s (RTT 3.0)

8: 'CPU-Core (SSE2/AES)': 604.4 PMKs/s (RTT 3.1)

CUDA:

1: 'CUDA-Device #1 'GeForce GTX 1080 Ti'': 161781.1 PMKs/s (RTT 0.8)

Hackitfree commented 6 years ago

@ PastiiJ ..I see you have 8 threads of processor working in pyrit and you get 160 000PMK/s..Its GTX 1080 Ti -You should get more...Disable all cores but 1 in .Pyrit / config files writing No of cpu's=1..You should get much better result...And free resources to do something else ;)..Anyway i get on my GTX 1060 128 000 PMK/s maximum speed...1 CPU core working..Tell me if it improved please :)

PastiiJay commented 6 years ago

Wooow thx Hackitfree !!

root@KaliDesktop:/opt/Pyrit-master/modules/cpyrit_cuda# pyrit benchmark Pyrit 0.5.1 (C) 2008-2011 Lukas Lueg - 2015 John Mora https://github.com/JPaulMora/Pyrit This code is distributed under the GNU General Public License v3+

Running benchmark (224931.1 PMKs/s)... |

Computed 224931.07 PMKs/s total.

1: 'CPU-Core (SSE2/AES)': 1341.9 PMKs/s (RTT 2.9)

CUDA:

1: 'CUDA-Device #1 'GeForce GTX 1080 Ti'': 244530.3 PMKs/s (RTT 0.5)

Hackitfree commented 6 years ago

How to Install Pyrit Cuda + Toolkit 8 + Nvidia GTX 1060 in Kali Linux 2017.3

*Clean Install Kali 2017.3 Kernel 4.13.0-kali1-amd64 and Update

apt update && apt upgrade && apt autoremove && apt -f install && apt upgrade

apt remove pyrit

echo -e "blacklist nouveau\noptions nouveau modeset=0\nalias nouveau off" > /etc/modprobe.d/blacklist-nouveau.conf

update-initramfs -u && reboot

*Verify nouveau is disabled

lsmod | grep -i nouveau

*must be a blank return

apt install nvidia-driver ocl-icd-libopencl1 nvidia-cuda-toolkit

apt install mesa-utils nvidia-cuda-gdb nvidia-cuda-doc libssl-dev zlib1g-dev libpcap-dev libgl1-nvidia-glx python2.7-dev python-pycuda-doc python-pycuda python3-pycuda

reboot

*Pyrit prep

git clone https://github.com/JPaulMora/Pyrit.git

gedit /root/.bashrc

*Add these lines to the end of .bashrc

PATH=$PATH:/usr/lib/nvidia-cuda-toolkit/bin. LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/nvidia-cuda-toolkit/lib:/lib export PATH export LD_LIBRARY_PATH

*Pyrit_Cuda Prep and Install

cd /root/Pyrit/modules/cpyrit_cuda/

gedit setup.py

Edit the setup.py, line 35,edit it from*

for path in ('/usr/local/cuda', '/opt/cuda'):

*to:

for path in ('/usr/local/cuda','/usr/lib/nvidia-cuda-toolkit','/opt/cuda'):

line 88,edit it from**

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang-3.8 -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

*then

cd /root/Pyrit

*and install

python setup.py clean build install

*last thing

gedit /root/.pyrit/config

*change from:

use_CUDA = false

*to

use_CUDA = true

*now check your GPU

pyrit list_cores

ENJOY !!!

Note: GPU works best with 1 CPU core gedit /root/.pyrit/config set: limit_ncpus = 1

My score:

root@kali:/# pyrit benchmark Pyrit 0.5.1 (C) 2008-2011 Lukas Lueg - 2015 John Mora https://github.com/JPaulMora/Pyrit This code is distributed under the GNU General Public License v3+

Running benchmark (120907.2 PMKs/s)... \

Computed 120907.24 PMKs/s total.

1: 'CPU-Core (SSE2/AES)': 1112.1 PMKs/s (RTT 3.1)

CUDA:

1: 'CUDA-Device #1 'GeForce GTX 1060 3GB'': 130434.6 PMKs/s (RTT 1.0)

kudzie567 commented 6 years ago

Thanks for the above instructions, now i can use my gpu successfully

PastiiJay commented 6 years ago

root@KaliDesktop:~# pyrit benchmark Pyrit 0.5.1 (C) 2008-2011 Lukas Lueg - 2015 John Mora https://github.com/JPaulMora/Pyrit This code is distributed under the GNU General Public License v3+

Running benchmark (239409.9 PMKs/s)... /

Computed 239409.86 PMKs/s total.

1: 'CPU-Core (SSE2/AES)': 1470.9 PMKs/s (RTT 2.9)

CUDA:

1: 'CUDA-Device #1 'GeForce GTX 1080 Ti'': 262496.3 PMKs/s (RTT 0.5)

Just OC first core of my 6700k to 4,7Ghz ≃ 18000PMKs/s more than stock (244530.3 PMKs/s) with limit_ncpus = 1

CPU bottleneck GPU, so i try to disable all cores: edit /root/.pyrit/config limit_ncpus = -4

But it crash after 30sec..

Can we run this only with GPU?

Thx

h3xcmd commented 6 years ago

NOTE TO SELF

cd /root/Pyrit/modules/cpyrit_cuda/ gedit setup.py

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang-3.8 -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu

@kudzie567 @Andiansyah23

pwnhun73r commented 6 years ago

@Hackitfree can you help me please, i have a laptop hp omen with intel and nvidia 1050 and i want install nvidia driver with cuda for use pyrit. is the same method that in your answer?

Hackitfree commented 6 years ago

@ c3ph3u5qwerty - Well i guess its same with your laptop-One thing..If your laptop uses the integrated graphics and switch between modes then you should consider this tricks with Optimus switching card:

https://forums.kali.org/showthread.php?35748-TUTORIAL-Installing-official-NVIDIA-driver-in-Optimus-laptop&highlight=nvidia http://www.hackdig.com/?07/hack-4577.htm https://svn.boost.org/trac10/attachment/ticket/11852/float128GccNvcc.patch

This can help.. I didnt try installing the newest version of KaliLinux yet-i hope something is improved with their installed already version of Pyrit..However try my guide and post what you have done..I will try resolving any issue you might have..

pwnhun73r commented 6 years ago

@Hackitfree very very thanks, this method works :) but i have a question with pyrit (i use parrot OS on other laptop, and installed pyrit compatible with opencl). and works but when reboot my laptop pyrit doesn't recognize my gpu. you know what's going on?

Hackitfree commented 6 years ago

@ pwnhun73r Hmmm.. So you made pyrit work-but when you restarted laptop-it wouldnt recognise GPU?..It might be some problem with switchable graphics in your laptop..Try to solve with some trick with "Optimus": https://forums.kali.org/showthread.php?35748-TUTORIAL-Installing-official-NVIDIA-driver-in-Optimus-laptop&highlight=nvidia http://www.hackdig.com/?07/hack-4577.htm https://svn.boost.org/trac10/attachment/ticket/11852/float128GccNvcc.patch

I hope it will help-i never worked with Pyrit on laptop swichable graphics so i dont know where is the problem...

kur4kwdz commented 6 years ago

Could u help me with this:

root@kali:/opt/Pyrit/modules/cpyrit_cuda# python setup.py clean build install running clean Removing temporary files and pre-built GPU-kernels... running build running build_ext Compiling CUDA module using nvcc 8.0, V8.0.61... Executing '/usr/local/cuda/bin/nvcc -m64 -ccbin clang-3.8 -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning). clang-3.8: No such file or directory Traceback (most recent call last): File "setup.py", line 166, in setup(**setup_args) File "/usr/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/usr/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/usr/lib/python2.7/distutils/command/build.py", line 128, in run self.run_command(cmd_name) File "/usr/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "setup.py", line 91, in run subprocess.check_call(nvcc_cmd, shell=True) File "/usr/lib/python2.7/subprocess.py", line 190, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '/usr/local/cuda/bin/nvcc -m64 -ccbin clang-3.8 -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu' returned non-zero exit status 1

Hackitfree commented 6 years ago

Did you follow all instructions? If yes then try to compile it without any changes in lines of setyp.py.. The newest version of Kali dont need this changes!

kur4kwdz commented 6 years ago

without any changes compiler returns: `root@kali:/opt/Pyrit/modules/cpyrit_cuda# python setup.py clean build install

running clean

Removing temporary files and pre-built GPU-kernels...

running build

running build_ext

Compiling CUDA module using nvcc 8.0, V8.0.61...

Executing '/usr/local/cuda/bin/nvcc -m64 -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).

nvcc fatal : The version ('40001') of the host compiler ('clang') is not supported

Traceback (most recent call last):

File "setup.py", line 166, in 

setup(**setup_args)

File "/usr/lib/python2.7/distutils/core.py", line 151, in setup

dist.run_commands()

File "/usr/lib/python2.7/distutils/dist.py", line 953, in run_commands

self.run_command(cmd)

File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command

cmd_obj.run()

File "/usr/lib/python2.7/distutils/command/build.py", line 128, in run

self.run_command(cmd_name)

File "/usr/lib/python2.7/distutils/cmd.py", line 326, in run_command

self.distribution.run_command(command)

File "/usr/lib/python2.7/distutils/dist.py", line 972, in run_command

cmd_obj.run()

File "setup.py", line 91, in run

subprocess.check_call(nvcc_cmd, shell=True)

File "/usr/lib/python2.7/subprocess.py", line 190, in check_call

raise CalledProcessError(retcode, cmd)

subprocess.CalledProcessError: Command '/usr/local/cuda/bin/nvcc -m64 -ccbin clang -Xcompiler

"-fPIC" --ptx ./_cpyrit_cudakernel.cu' returned non-zero exit status 1 `

Hackitfree commented 6 years ago

So you followed my tutorial ?? Try compiling like this with clang 3.8:

*Pyrit_Cuda Prep and Install

cd /root/Pyrit/modules/cpyrit_cuda/

gedit setup.py

Edit the setup.py, line 35,edit it from*

for path in ('/usr/local/cuda', '/opt/cuda'):

*to:

for path in ('/usr/local/cuda','/usr/lib/nvidia-cuda-toolkit','/opt/cuda'):

line 88,edit it from**

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

nvcc_cmd = NVCC + bit_flag + ' -ccbin clang-3.8 -Xcompiler "-fPIC" --ptx ./_cpyrit_cudakernel.cu'

*then

cd /root/Pyrit

*and install

python setup.py clean build install

*last thing

gedit /root/.pyrit/config

*change from:

use_CUDA = false

*to

use_CUDA = true

*now check your GPU

pyrit list_cores

You must have some kind of "clang" installed.. Did you install Nvidia cuda? Tell me what result will get now!

Hackitfree commented 6 years ago

Try this also:

cd /root/Pyrit/modules/cpyrit_cuda

python setup.py clean build install

cd /root/Pyrit/

python setup.py clean build install

yes 2 times in differnt dir

vido89 commented 6 years ago

@Hackitfree Tnx for instructions I followed it but Im getting an error. Im on ubuntu 16.04 driver 384.130 and cuda 10


slobodan@G31M-ES2L:~/Pyrit/modules/cpyrit_cuda$ pyrit list_coresPyrit 0.5.1 (C) 2008-2011 Lukas Lueg - 2015 John Mora
https://github.com/JPaulMora/Pyrit
This code is distributed under the GNU General Public License v3+

Traceback (most recent call last):
  File "/usr/local/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.7/dist-packages/pyrit_cli.py", line 117, in initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.7/dist-packages/pyrit_cli.py", line 294, in list_cores
    with cpyrit.cpyrit.CPyrit() as cp:
  File "/usr/local/lib/python2.7/dist-packages/cpyrit/cpyrit.py", line 442, in __init__
    self.CUDAs.append(CUDACore(queue=self, dev_idx=dev_idx))
  File "/usr/local/lib/python2.7/dist-packages/cpyrit/cpyrit.py", line 243, in __init__
    _cpyrit_cuda.CUDADevice.__init__(self, dev_idx)
SystemError: Unknown CUresult.