In this sense I thought it's maybe not spamming to give some guide how I could setup the poclbm miner from the beginning on an fresh ubuntu when using an amd gpu.
That maybe not a new future by Ubuntu, but as a windows user I was absolutily impressed that one can directly install and also uninstall ubuntu in windows on a windows partition.
I am using Ubuntu 12.04 64bit and a HD 5850:
Install AMD-catalyst for linux 64bit:
1) get AMD-catalyst, extract, and go to extraction folder
2) "sudo sh ./AMD-<blablabla>.run"
3) restart: "sudo shutdown 0 -r"
Install AMD-APP-SDK:
1) get AMD-APP-SDK, extract and go to extraction folder
2) "sudo sh ./Install-AMD-SDK.sh"
3) restart: "sudo shutdown 0 -r"
Packages I installed making pyopencl "make" running (compilation) using
"sudo apt-get install <packages>":
1) python-mako (I do not know if this is really needed)
2) python-setuptools
3) ipython
4) python-numpy (I do not remember, if this was already installed)
5) g++
6) libboost-all-dev (I could not install this packs in the first place, but after installing "aptitude" to find the problem, the problem was solved)
7) subversion (I do not know if this is really needed)
download and extract pyopencl from "
https://pypi.python.org/pypi/pyopencl", go to extraction folder then:
1) "python configure.py --cl-inc-dir=/opt/AMDAPP/include --cl-lib-dir=/opt/AMDAPP/lib/x86_64 --cl-libname=OpenCL"
2) "make"
3) " sudo make install"
get poclbm-skc source and compile skeinhash lib:
1) "git clone
https://github.com/snoopcode/poclbm-skc.git"
2) "cd poclbm-skc\skeinhash"
3) "gcc -fPIC -shared -o skeinhash.so *.c"
go back to poclbm main folder and check funtionality of poclbm:
1) "ipython"
2) "run poclbm"
3) The output shows your GPU devices ... maybe you have different devices on diffrent platforms try: "run poclbm --platform=<platformID>"
4) restart ipython: "exit()"-> "ipython" (if I don't do that I get an error)
5) "run poclbm --platform=<platformID> --device=<deviceID> stratum://<login>.<workername>:<passwort>@skc.coinmine.pl:6400"