I want to implement a sandbox in GPU kernel to protect the memory content in a specific memory range. Thus, I wonder are there methods to control the virtual address range of specific objects within a GPU kernel? For comparison, in Arm, I can use a link script to determine the starting virtual address for each section in the binary.
this would probably be the closest thing
I’m not suggesting its an exact duplicate of whatever you are doing with a link script. It does not give you control over sections in a binary.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.