In cuda-gdb, is it possible to set a hardware watchpoint on a variable in global memory? I’ve tried commands like this:
watch (@global float) var
but they seem to have no effect.
I’m using Fermi cards and am running
x86_64 Red Hat Enterprise Linux Client release 5.4 (Tikanga)
Nvidia driver version 256.40
The Cuda toolkit I downloaded was cudatoolkit_3.1_linux_64_rhel5.4.run