Can ST Cube IDE do 32F4 execution time profiling?

The internet is full of posts from people tearing their hair out
trying to use this tool :)

I have it running OK although 99% of it is a mystery. And the code
which the configurator generates is crap ;)

With the old ICEs one could do a bar graph easily, across the address
range, and then you could zoom in.

Obviously there is no identical way to do this because the 32F runs
code fully internally, so the only way is with a temporary breakpoint,
cancelling it as fast as the debugger is capable of (probably still
takes milliseconds) and moving it to the next address, the next C
source line probably.