Ruyman
Ruyman
> @jeffhammond but it's not really a SYCL issue - it's a DPC++ issue. hipSYCL and presumably ComputeCpp PTX backend will not suffer from that. I think this PR replaces...
Thanks but I would probably find it easier to contribute something smarter to the DPC++ CUDA backend than get you some test code used in our implementation! :-)
Just wrote a quick patch, https://github.com/intel/llvm/pull/2724, can someone test it on a decent NVIDIA gpu?
@ProGTX can you also remove the usage of c++14 lambdas from the test before merging?
Seems like a reasonable restriction to me. My concern is that we cannot assume all implementations of execution resource / context would allow for this fine-grained selection of resources. 1....
I think it makes sense. I would like to see something like a memory_context equivalent of the execution_context, but that would not be ready in time for the next meeting.
To me it seems that the executor context should be some sort of interface that implementations can use to "plug-in" into the executor context interface, so I lean towards A....
I am changing the label to enhancement since this will not be addressed in the paper for RAP18
I am removing the RAP18 label since this is something we can address after the meeting