Pages:
Author

Topic: python OpenCL bitcoin miner (Read 1239060 times)

newbie
Activity: 10
Merit: 0
February 25, 2014, 05:15:45 PM
Nice one man, thanks for the help!
sr. member
Activity: 266
Merit: 250
December 31, 2013, 07:58:01 AM
Is this better than GUIminer?
I think so.
member
Activity: 89
Merit: 10
June 11, 2013, 05:46:11 AM
The miner check the result of the opencl kernel by doing the double sha calculation with the nonce just found by the kernel.

Makes sense.
Thank you.
full member
Activity: 193
Merit: 100
June 10, 2013, 05:19:04 PM
The miner check the result of the opencl kernel by doing the double sha calculation with the nonce just found by the kernel.
member
Activity: 89
Merit: 10
June 10, 2013, 04:22:45 PM
Hello fellow miners.

I use poclbm without any major problems, but every now and then I will get an error message:
Code:
Verification failed, check hardware! (, <32bit hex value>)

The general advice is to check one's hardware and I concur, but I want to look at it from a different angle.

My question is on the mathematical side. What exactly is tested to perform this verification?
I haven't seen it described anywhere.

full member
Activity: 193
Merit: 100
May 31, 2013, 08:06:13 AM
A possible workarround for people that have hardware error with the driver 13.4 and can't downgrade the driver ... :

The BFI_INT optimisation look like to be problematic on that driver version, more informations :

https://bitcointalk.org/index.php?topic=221041.new#new

EDIT : i think it should be a good idea to define or not BFI_INT depending on the version of the driver also and not just only the plateform name (aka Jupiter for my 5770)
member
Activity: 70
Merit: 10
May 13, 2013, 12:07:36 PM
Is this better than GUIminer?
newbie
Activity: 14
Merit: 0
May 11, 2013, 04:28:46 AM
what operating system do you recommend?
full member
Activity: 171
Merit: 127
May 07, 2013, 09:46:19 AM
@Ruudsy

Just put the server definition after all other options
full member
Activity: 193
Merit: 100
May 07, 2013, 08:55:39 AM
At least one server is required

That's the important error message, you didn't provide any mining pool server address.
newbie
Activity: 19
Merit: 0
May 07, 2013, 07:49:55 AM
Tried that before, is not working unfortunatelly ...

Code:
sudo fglrxinfo

gives me this

Code:
display: :0  screen: 0
OpenGL vendor string: Advanced Micro Devices, Inc.
OpenGL renderer string: ATI Radeon HD 5800 Series
OpenGL version string: 4.2.12217 Compatibility Profile Context 12.104

so I assume a
Code:
export DISPLAY=:0.0
is required? Or do I need
Code:
export DISPLAY=:0
?

Neither ist working ...

Here is everything in one piece:
Code:
xxx@Eibe:/opt/poclbm$ export DISPLAY=:0
xxx@Eibe:/opt/poclbm$ sudo python poclbm.py

WARNING: no adl3 module found (github.com/mjmvisser/adl3), temperature control is disabled


OpenCL devices:

[0] Cypress
[1] Intel(R) Core(TM)2 Duo CPU     E6750  @ 2.66GHz

No devices specified, using all GPU devices


At least one server is required

xxx@Eibe:/opt/poclbm$

legendary
Activity: 3920
Merit: 2349
Eadem mutata resurgo
May 07, 2013, 06:05:25 AM
Hey,

i switched from Windows to Ubuntu 12.04 and don't get poclbm running.

I installed ATI Drivers, 12.104 Catalyst as well as SDK 2.8, pyopencl is also running perfectly. But I don't get poclbm working.

The Error reaccurring all the time trying to run
Code:
sudo python poclbm.py http://xxxxxxxx:[email protected]:8332 -d0 -v -w128

The terminal always says

Code:
07/05/2013 12:28:18, Ignored invalid server entry: http://xxxxxxxx:[email protected]:8332


WARNING: no adl3 module found (github.com/mjmvisser/adl3), temperature control is disabled


At least one server is required


Sounds like you need to allocate the correct DISPLAY(s) (the ones associated with your GPU(s) ) to the Xserver ... it's been so long i forget how to do it ... something like

Code:
export DISPLAY = 0.Y
... where Y is number associated with your GPU
newbie
Activity: 19
Merit: 0
May 07, 2013, 05:34:11 AM
Hey,

i switched from Windows to Ubuntu 12.04 and don't get poclbm running.

I installed ATI Drivers, 12.104 Catalyst as well as SDK 2.8, pyopencl is also running perfectly. But I don't get poclbm working.

The Error reaccurring all the time trying to run
Code:
sudo python poclbm.py http://xxxxxxxx:[email protected]:8332 -d0 -v -w128

The terminal always says

Code:
07/05/2013 12:28:18, Ignored invalid server entry: http://xxxxxxxx:[email protected]:8332


WARNING: no adl3 module found (github.com/mjmvisser/adl3), temperature control is disabled


At least one server is required

member
Activity: 73
Merit: 10
April 10, 2013, 09:35:44 PM
I'd like a Windows build as well.

I use my own front-end:
https://bitcointalksearch.org/topic/crypto-miners-in-tray-a-lightweight-front-end-for-unattended-mining-149442
http://stanislavs.org/bitcoin-miners-in-tray-miner-gui-manager/
, but I am forced to have GUIMiner for the sake of the bundled poclbm build, which supports HD 7970.

poclbm_py2exe_20120920.7z linked to in the opening comment gives the following error:

09.04.2013 20:54:27 [AMD HD7970]: Traceback (most recent call last):
09.04.2013 20:54:27 [AMD HD7970]: File "poclbm.py", line 85, in
09.04.2013 20:54:27 [AMD HD7970]: File "BitcoinMiner.pyo", line 50, in __init__
09.04.2013 20:54:27 [AMD HD7970]: File "BitcoinMiner.pyo", line 263, in get_adapter_info
09.04.2013 20:54:27 [AMD HD7970]: NameError: global name 'ADLError' is not defined

The build, included with GUIMiner does not give any errors with the same parameters. I saw on github there were a few bugfix commits.

Can we have the latest and greatest build for Windows, please?

I get the same errors...
I second this, and would love to have an updated standalone package...at least if there is I can't find it Smiley
legendary
Activity: 1680
Merit: 1014
April 09, 2013, 02:03:36 PM
I'd like a Windows build as well.

I use my own front-end:
https://bitcointalksearch.org/topic/crypto-miners-in-tray-a-lightweight-front-end-for-unattended-mining-149442
http://stanislavs.org/bitcoin-miners-in-tray-miner-gui-manager/
, but I am forced to have GUIMiner for the sake of the bundled poclbm build, which supports HD 7970.

poclbm_py2exe_20120920.7z linked to in the opening comment gives the following error:

09.04.2013 20:54:27 [AMD HD7970]: Traceback (most recent call last):
09.04.2013 20:54:27 [AMD HD7970]: File "poclbm.py", line 85, in
09.04.2013 20:54:27 [AMD HD7970]: File "BitcoinMiner.pyo", line 50, in __init__
09.04.2013 20:54:27 [AMD HD7970]: File "BitcoinMiner.pyo", line 263, in get_adapter_info
09.04.2013 20:54:27 [AMD HD7970]: NameError: global name 'ADLError' is not defined

The build, included with GUIMiner does not give any errors with the same parameters. I saw on github there were a few bugfix commits.

Can we have the latest and greatest build for Windows, please?
sr. member
Activity: 574
Merit: 250
April 09, 2013, 12:48:46 PM
Yes, I still maintain the miner. It is now modular, i.e. you can implement your own hardware or work-source types. I just haven't released a py2exe recently (since BFL support actually) - POCLBM doesn't look so popular anymore and works fine for anyone able to setup python + pip and pull sources from github.

I know a few people that run poclbm in the background on windows mostly just to feel they are supporting an alternative to fiat.   They don't follow the scene at all  though. I just let them know when I see an update to the windows build to let them know to update.  We have been figuring that any worthwhile  updates would get check pointed by a build.  Is this not really true?  Have windows build updates ceased then?
full member
Activity: 193
Merit: 100
April 09, 2013, 10:22:44 AM
Look like the setup.py isn't really installable

running 'python setup.py install' install only poclbm.py and not other py files, so it can't start.

Did you need help to create a correct setup for linux ? or you care only of windows py2exe release ?
newbie
Activity: 47
Merit: 0
March 31, 2013, 07:56:25 PM
Hi guys, I've been at this all day and I'm struggling to get poclbm to work. Is fglrx really the issue or is it something else? Any help would be greatly appreciated.

Code:
[ec2-user@ip-xx-xx-xx-xxx poclbm]$ sudo python ./poclbm.py http://xxxxx.xxx:[email protected]:3333 -v -w 256 --verbose -d 0 --platform 0
FATAL: Module fglrx not found.
Error! Fail to load fglrx kernel module! Maybe you can switch to root user to load kernel module directly

WARNING: no adl3 module found (github.com/mjmvisser/adl3), temperature control is disabled

 01/04/2013 00:50:09, started OpenCL miner on platform 0, device 0 (Tesla M2050)
 01/04/2013 00:50:09, Setting server (xxxxx.xxx @ stratum.bitcoin.cz:3333)
Traceback (most recent call last):
  File "./poclbm.py", line 101, in
    if switch: switch.stop()
  File "/home/ec2-user/poclbm/Switch.py", line 149, in stop
    self.server_source().stop()
  File "/home/ec2-user/poclbm/Switch.py", line 267, in server_source
    import GetworkSource
  File "/home/ec2-user/poclbm/GetworkSource.py", line 3, in
    from httplib import HTTPException
  File "/usr/lib64/python2.6/httplib.py", line 1094, in
    import ssl
  File "/usr/lib64/python2.6/ssl.py", line 83, in
    class SSLSocket(socket):
TypeError: Error when calling the metaclass bases
    function() argument 1 must be code, not str

OpenCL info script

Code:
[ec2-user@ip-xx-xx-xx-xxx ~]$ sudo python ./opencl.py
[01:48:38] Python OpenCL Info v0.1
[01:48:38] Python Version: 2.6.8 64bit
[01:48:38] PyOpenCL Path: /usr/lib64/python2.6/site-packages/pyopencl-2013.1-py2.6-linux-x86_64.egg/pyopencl
[01:48:38] Boost Python Version: Not Found
[01:48:39] PyOpenCL Version: 2013.1
FATAL: Module fglrx not found.
Error! Fail to load fglrx kernel module! Maybe you can switch to root user to load kernel module directly
[01:48:43] Listing platforms and devices:
[01:48:43]
[01:48:43] [cl:0] NVIDIA CUDA
[01:48:43]        [cl:0:0] Tesla M2050
[01:48:43]        [cl:0:1] Tesla M2050
[01:48:43]
[01:48:43] [cl:1] AMD Accelerated Parallel Processing
[01:48:43]        [cl:1:0] Intel(R) Xeon(R) CPU           X5570  @ 2.93GHz
[01:48:43]
[01:48:43] This program will exit in 300 seconds...

$ rpm -qa *\nvidia\*  *\kernel\*|sort;uname -r;lsmod |grep -e nvidia -e nouveau;cat /etc/X11/xorg.conf

Code:
kernel-2.6.35.11-83.9.amzn1.x86_64
kernel-3.4.37-40.44.amzn1.x86_64
kernel-headers-3.4.37-40.44.amzn1.x86_64
nvidia-310.40-2013.03.10.amzn1.x86_64
nvidia-kmod-2.6.35.11-83.9.amzn1-260.19.36-1.27.amzn1.x86_64
nvidia-kmod-3.4.37-40.44.amzn1-310.40-2013.03.10.amzn1.x86_64
3.4.37-40.44.amzn1.x86_64
nvidia               9399284  0
i2c_core               24483  2 nvidia,i2c_piix4
cat: /etc/X11/xorg.conf: No such file or directory
legendary
Activity: 1001
Merit: 1005
January 26, 2013, 04:16:35 AM
my fault, it's work Grin
legendary
Activity: 1001
Merit: 1005
January 26, 2013, 04:07:34 AM
added option --no-bfl , but get new problem
Code:
arm@rig-1:~/mining/poclbm$ ./run001.sh

WARNING: no adl3 module found (github.com/mjmvisser/adl3), temperature control is disabled

 26/01/2013 13:03:34, Setting server (xxx @ api.bitcoin.cz:8332)
api.bitcoin.cz:8332 26/01/2013 13:03:34, checking for stratum...
api.bitcoin.cz:8332 26/01/2013 13:03:34, started OpenCL miner on platform 0, device 0 (Cypress)
./run001.sh: cтpoкa (string) 1:  2408 Oшибкa ceгмeнтиpoвaния  (segmentation fault)                   (core dumped) python ./poclbm.py http://xxx:[email protected]:8332 -f 30 -v -w 256 -d 0 --platform=0 --no-bfl
Pages:
Jump to: