How can I measure texture hit and miss with a cudaprof config file on C1060

Accelerated Computing CUDA CUDA Programming and Performance
1

Apparently it is possible to measure texture hit and miss on a C1060 (compute capability 1.3) but I don’t know what the codes are for the counters.

I’m using cudaprof in linux, setting CUDAPROF=1, declaring a config file inidicating which counters to use, using such codes as divergent_branch or gld_coherent. I would like to measure texture hit and miss, but I don’t know what the codes are for these.

What are they?