Suppose I have a function presented in LLVM IR, and I want to estimate the number of cpu cycles that this function will execute on my machine.
I know that this is architecture-specific information, but I want to get a rough estimate of this for some planning purposes.
Is it possible to write a skip that iterates over instructions in a function and returns the total number of loops that these instructions will execute?
Would thank for any ideas.
Thanks,
Kayhan
source
share