python OpenCL bitcoin miner

[willebra]

Anybody know how to fix this error, or what is causing it??

Traceback (most recent call last):
  File "poclbm.py", line 27, in <module>
    platforms = cl.get_platforms()
pyopencl.LogicError: clGetPlatformIDs failed: invalid/unknown error code

I get this same error, when using poclbm.py under linuxcoin

Answering myself: it seems I had not accepted the AMD license agreement, and it seemed to solve the question.

[malitas09]

What does it look like when you receive bitcoins from mining, does it display anything on the screen that shows the hash rate? Or does it just showup as a transaction in bitcoin?

[VVV]

Hello.

I have ATI 4670.
I use poclbm_py2exe_20110428.
But when I start poclbm.exe it show only my CPU.
ATI APP SDK already installed(x64 for Win7 x64),
Here is GPU Caps Viewer
http://s1.ipicture.ru/uploads/20110609/353mu6k3.jpg
On key press "OpenCL demos -> Start"  got error "OpenCL GPU not supported on the selected platform. Bye..."

Why? 4670 don't support OpenCL?

[willebra]

What does it look like when you receive bitcoins from mining, does it display anything on the screen that shows the hash rate? Or does it just showup as a transaction in bitcoin?

You will normally be mining with a mining pool. You can see your shares at the pools website and there will your bitcoins grow to your account. You can then transfer them to your bitcoin address via the pool's website.

[sturle]

I'm getting lots of "warning: job finished, miner is idle".

I have the same problem on Linux, but only when using -v (vectors) and a low value for -f.  I never see it if I don't use -v, and I never see it with default -f.  Haven't tested for very long with higher -f, but at 10 the messages are mostly gone.  I mine solo, of course, and getwork calls takes neglible time.  I made each call to getwork log a line, and the error shows up when it takes > 12 seconds between two getwork calls.  My askrate is 5 seconds, but there is usually at least six seconds between each call to getwork.  Sometimes much more.  A lower value for -a does not remove the long gaps between getwork calls which are followed by "miner is idle" messages.

[Moussekateer]

Any chance of a compiled exe version of the new poclbm with backup pools?

[d3m0n1q_733rz]

So, it would appear that Malware-bytes has listed Poclbm as a trojan.  I've taken the liberty of sending in a false positive report already.  If anyone wants to back me up, the post is here http://forums.malwarebytes.org/index.php?showtopic=87882.  It'll also keep you up-to-date on what's going on with them.  Should they deny the request, feel free to ask them again yourselves since I know many of these anti-malware companies get a little headstrong and need some convincing.

[d3m0n1q_733rz]

As an update, they have added poclbm as a false positive in their next update; so anyone using Malwarebytes will not accidentally remove their bitcoin miner.

[btcmonkey]

Is target difficulty of 1 hardcoded into this client?  I'm experimenting with pools that vary share difficulty and am getting unexpected results.

[Zenitur]

I found an error in newest GIT (June 17, 2011).

Code:

deepbit.net:8332 23/06/2011 19:01:15, 8617ca99, accepted
deepbit.net:8332 23/06/2011 19:02:50, 665986ba, accepted
deepbit.net:8332 23/06/2011 19:05:20, long poll: new block 000003d711afefc1
deepbit.net:8332 23/06/2011 19:05:27, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:05:27, LP connected to deepbit.net:8332
deepbit.net:8332 23/06/2011 19:05:37, long poll: new block 0000110787893cdb
deepbit.net:8332 23/06/2011 19:05:43, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:05:43, LP connected to deepbit.net:8332
deepbit.net:8332 23/06/2011 19:07:22, long poll: new block 0000065c454db6b4
deepbit.net:8332 23/06/2011 19:07:28, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:07:28, LP connected to deepbit.net:8332
deepbit.net:8332 23/06/2011 19:08:56, long poll: new block 000004edee027877
deepbit.net:8332 23/06/2011 19:09:02, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:09:02, LP connected to deepbit.net:8332
deepbit.net:8332 23/06/2011 19:11:37, d3627bf3, accepted
deepbit.net:8332 23/06/2011 19:12:30, long poll: new block 00001214acf5b4c8
deepbit.net:8332 23/06/2011 19:12:36, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:12:36, LP connected to deepbit.net:8332
deepbit.net:8332 23/06/2011 19:12:48, 4b7aea48, accepted
deepbit.net:8332 23/06/2011 19:13:30, 26100840, accepted
deepbit.net:8332 23/06/2011 19:13:44, 4196b69e, accepted
deepbit.net:8332 23/06/2011 19:14:35, c64e9b09, accepted
deepbit.net:8332 23/06/2011 19:14:48, 66bc4256, accepted
deepbit.net:8332 23/06/2011 19:20:00, long poll: new block 00000df42b538904
deepbit.net:8332 23/06/2011 19:20:01, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:20:01, LP connected to deepbit.net:8332
deepbit.net:8332 23/06/2011 19:23:46, long poll: new block 0000009efde53e3e
deepbit.net:8332 23/06/2011 19:23:53, Using new LP URL /listenChannel
deepbit.net:8332 23/06/2011 19:23:53, LP connected to deepbit.net:8332
deepbit.net:8332 31526 khash/s^C

[RaTTuS]

I found an error in newest GIT (June 17, 2011).

which error

[m0mchil]

New version is up. It incorporates improvements from many developers.

Kudos go to:

Phateus - phatk kernel 124RraPqYcEpX5qFcQ2ZBVD9MqUamfyQnv

fpgaminer - Using poclbm With The phatk Kernel 1NT4RyJMqtRuDRr6zHdXdKSpmX3SR5he6z

bitless, rethaw - 3% faster mining 15igh5HkCXwvvan4aiPYSYZwJZbHxGBYwB

Kyle Gibson - Fallback mining pool(s) 1BdWJEUCkm6on231iX2ThXZHXZDfQny69R

laanwj - sleep per frame

[fasti]

For me 275 gtx mining, from previous 20110428 to current 20110627 -> -12% performance.

[Soak]

New release (2011-06-27) : https://github.com/downloads/m0mchil/poclbm/poclbm_py2exe_20110627.7z

Added an improvement in kernel (use phatk with modification of Ma formula).
+3% hashrate in average.

[theowalpott]

also got a performance hit on nvidia.

[zimpixa]

This means new version no longer uses old kernel? Or is there any option to select it? (If no please add one, cuz Im getting better results with version one on one card and better with old version on other one AND I still want to use guiminer)

[dishwara]

This means new version no longer uses old kernel? Or is there any option to select it? (If no please add one, cuz Im getting better results with version one on one card and better with old version on other one AND I still want to use guiminer)

In GUIminer folder, go to kernels ->phatk & in kernel.cl edit the line as said in other post.
Also, you can edit, kernels -> poclbm , kernel.cl

No need to download or wait for new GUIminer for this performance increase.

[Mobius]

Gigabyte 5870 875gpu,300mem,1.140v went from 384 to 399
Gigabyte 5830 875gpu,333mem,1.140v went from 264 to 280
Gigabyte 5830 899gpu,333mem,1.140v went from 274 to 287

I liked the statistical displays of the FPGAMINER, Luke-Jr and tweeks by ULEE

Thanks to all of you-Great Work

[Luke-Jr]

New version is up.

It doesn't work:

Code:

mining.eligius.st:8337 28/06/2011 00:53:49, Setting pool 1GEJfZRPrK2BLSSx3r6gwtuFxCUvq3QytN @ mining.eligius.st:8337
Killed

It incorporates improvements from many developers.

You forgot mine.

Phateus - phatk kernel 124RraPqYcEpX5qFcQ2ZBVD9MqUamfyQnv

This is probably the problem. Phoenix with phatk also Killed-crashes.

laanwj - sleep per frame

What does this actually do?

[twmz]

New release (2011-06-27) : https://github.com/downloads/m0mchil/poclbm/poclbm_py2exe_20110627.7z

Added an improvement in kernel (use phatk with modification of Ma formula).
+3% hashrate in average.

Can you re-add the old kernel as an optional kernel?  My mac doesn't appear to work with the phatk kernel (lots of failed validation messages).

For now, I just am using the previous revision from SVN, but would like to continue to receive other future improvements and don't really want to maintain my own fork.

Thanks!