I have just looked at the nVidia example that covers the nv_var_mem_mgr from their SDK. Of course, this was after I wrote a memory manager myself. However, I am confused as to how it can work after looking at the code for it. Is there a description of it’s usage and operations somewhere. To me it looks like there could be lots of problems with the method in which memory is used. In particular, allocating sections of memory which have not been fenced.
I have looked around, but I have found nothing but this new code example which doesn’t really say much about the usage.