Re: KHTML-Measurement (from my talk at aKademy)

On Thursday August 26 2004 18:04, Josef Weidendorfer wrote:
> > > As I said, I plan to add a heuristic to KCachegrind to extract the
> > > callgraph from a simulation run, and use this with sampling data from
> > > OProfile to get inclusive costs.
> >
> > Oh, that would be a dream come true! Please please please do it!
>
> Wow. You think this could be a killer feature for KCachegrind?

Absolutely! If you've ever used Intel's VTune, that's exactly how it works. It 
has two ways of profiling a program: one using sampling (using performance 
counters or the OS timer), and the other instruments the functions so that it 
generates a callgraph. Then it blends the two sources of information to 
present a unified report.

> Currently I am improving support for more general formulas for
> inherited event types (e.g. "FLOP / Time" to be able to show flop rates).
> I hope to come to a point where KCachegrind can be used as test coverage
> visualization, too.

Woah, very cool.

> But if there is no misprediction in a function with lots of branches, there
> is no penalty because of these branches. And I do not think that adding a
> simple branch prediction will have much overhead, as I already have data
> structures for each branch, and storing an additional "most probable
> target" to every branch site, with a small saturating counter, should be
> easy.
>
> Anyway, actually I am not sure if it's worth to improve simulation at
> all...

Sure, you're right, but I still think playing the probabilities is more 
important and, in the end, you shouldn't worry about branch prediction.

-- 
Simon Perreault <[email protected]> -- http://nomis80.org