Hi all,
I’m the author of gvirtus [1], a software component that enables virtual machines to use cuda application, running kernels on the host’s gpus.
In the new release of the cuda toolkit there are the two routine in the subject ({cuda,cu}GetExortTable), they are undocumented and the storage of their output values are unknown (there is just a pointer to a void pointer as output value). Of course for supporting the toolkit 3.2 in gvirtus i need to know how to handle these functions. So it’s strongly appreciated if someone (maybe from nvidia) can suggest me how to handle these routines.
Hi all,
I’m the author of gvirtus [1], a software component that enables virtual machines to use cuda application, running kernels on the host’s gpus.
In the new release of the cuda toolkit there are the two routine in the subject ({cuda,cu}GetExortTable), they are undocumented and the storage of their output values are unknown (there is just a pointer to a void pointer as output value). Of course for supporting the toolkit 3.2 in gvirtus i need to know how to handle these functions. So it’s strongly appreciated if someone (maybe from nvidia) can suggest me how to handle these routines.