We now "heat color" the instruction-level samples so you can get a nice visual picture of where time is spent in functions. :^)
When you select a function in the profile tree, we will now display a per-instruction breakdown of aggregated samples in that function. This allows us to look much closer at what our code is doing! :^)