Diablo-D3
commented
12 years ago Yes, OSX currently has a bug where completely valid work sizes get the error equivalent to "too large". Just add -w 64 to your arguments and it fixes it.
Closed ghost closed 12 years ago
Diablo-D3
commented
12 years ago Yes, OSX currently has a bug where completely valid work sizes get the error equivalent to "too large". Just add -w 64 to your arguments and it fixes it.
I am running 10.7.2 This is the error I'm getting:
[1/16/12 7:06:21 PM] Started
[1/16/12 7:06:21 PM] Connecting to: http://mining.bitcoin.cz:8332/
[1/16/12 7:06:21 PM] Using Apple OpenCL 1.1 (Jul 25 2011 15:56:07)
[1/16/12 7:06:33 PM] Added ATI Radeon HD 6750M (#1) (6 CU, local work size of 256) [1/16/12 7:06:34 PM] ERROR: [CL_INVALID_WORK_GROUP_SIZE] : OpenCL Error : clEnqueueNDRangeKernel failed: local_size0 != required local_size0 [1/16/12 7:06:34 PM] ERROR: Failed to queue kernel, error -54
I know that others have asked about this same issue and the solution seems to be to lower the -w to 64 or 128 but I do not know how to do this. I feel foolish asking because it's probably a very simple question, but I would greatly appreciate any instructions anyone might be able to give for changing the work size.