https://community.intel.com/t5/OpenCL-for-CPU/Possible-bug-with-atomic-cmpxchg/m-p/983137#M2604 <P>Sorry, I've just found that it is numerical error.</P> <P>Thanks.</P> Thu, 06 Jun 2013 11:09:57 GMT Ali_R_1 2013-06-06T11:09:57Z https://community.intel.com/t5/OpenCL-for-CPU/Possible-bug-with-atomic-cmpxchg/m-p/983136#M2603 <P>I wrote an OpenCL code to multiply a row vector into a Compressed Sparse Row matrix. But it gives me a different answer eachtime I run it.</P> <P>I have built a small repro case based on my matrcies. As one can see, the program breaks at differnet values of j, despite it is expected to print Success. I think the problem is related to&nbsp;atomic_cmpxchg cache flushing, since the loop containing it always run only one time, which is a little stange.</P> <P>Can any body help me on this please?</P> <P>Thanks.</P> Thu, 06 Jun 2013 10:26:38 GMT https://community.intel.com/t5/OpenCL-for-CPU/Possible-bug-with-atomic-cmpxchg/m-p/983136#M2603 Ali_R_1 2013-06-06T10:26:38Z https://community.intel.com/t5/OpenCL-for-CPU/Possible-bug-with-atomic-cmpxchg/m-p/983137#M2604 <P>Sorry, I've just found that it is numerical error.</P> <P>Thanks.</P> Thu, 06 Jun 2013 11:09:57 GMT https://community.intel.com/t5/OpenCL-for-CPU/Possible-bug-with-atomic-cmpxchg/m-p/983137#M2604 Ali_R_1 2013-06-06T11:09:57Z