I've seen behaviour similar to this being accepted by implementations, but you are correct in that the specification doesn't explicitly say that it is acceptable behaviour!
I've posted in the OpenCL forums here [[ http://www.khronos.org/message_boards/showthread.php/9301-Clarification-on-event_t-usage-in-OpenCL-specification | Clarification-on-event_t-usage-in-OpenCL-specification ]] to see if I can get some clarification. Thanks again for your help! http://llvm-reviews.chandlerc.com/D2860 _______________________________________________ cfe-commits mailing list cfe-commits@cs.uiuc.edu http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
