<#part sign=pgpmime> On Fri, 9 Mar 2012 15:37:27 -0800, "Eli Stevens (Gmail)" <[email protected]> wrote: > Thanks; I'd seen the note about --cuda-trace in old release notes > while ago, but since forgotten about it entirely. Cool. > > Here's what I'm seeing: > > cuParamSetv (mykernel) > cuParamSetSize (mykernel) > cuMemcpyHtoD > cuMemcpyHtoD > cuMemcpyHtoD > cuMemcpyHtoD > cuMemcpyHtoD > cuMemcpyHtoD > cuMemcpyHtoD > cuParamSetTexRef (mykernel) > cuParamSetTexRef (mykernel) > cuLaunchGrid (mykernel) > cuCtxSynchronize > > And then the active threads are: > > a 1 <_MainThread(MainThread, started 140345423361824)> > a 1 <CudaThread(Thread-25, started 140345099409152)> > ... > > Does this mean that the cuCtxSynchronize function was *called* or that > it *returned*? I'm assuming that it was called and is now blocking on > the kernel, but I want to make sure before I dive in.
This gets printed on entry to cuCtxSynchronize. You can easily hack the macros at the top of src/cpp/cuda.hpp to also log successful completion. Andreas _______________________________________________ PyCUDA mailing list [email protected] http://lists.tiker.net/listinfo/pycuda
