Re: timing information with perf

[Arnaldo Carvalho de Melo <acme@kernel.org>]

Em Mon, Nov 16, 2015 at 07:30:20PM -0500, Frank Ch. Eigler escreveu:
> acme wrote:
 
> > [...]
> > > For completeness, a complete systemtap implementation of this could look like:
> >  
> > > # stap -e '
> > >    global s, t%
> > >    probe kernel.trace("foobar_enter") {
> > >        t[tid()]=gettimeofday_us() 
> > >    } [...]
> > > '

> > Is that done syncronously?

> Meaning - calculated live as the tracepoints hit?  Yes, right during
> the respective callbacks.  No trace files nor postprocessing.

Right, 'perf trace' is in between those extremes, i.e. no synchronous
processing, no trace file for postprocessing. Just the ring buffer, that
is continuously processed.

- Arnaldo