PAPI in the ports

Harald Servat redcrash at
Thu Jun 14 10:20:35 UTC 2007


2007/6/14, Thomas Vogt <freebsdlists at>:
> Hi
> Thats sounds nice. You wrote "The goal of the PmcTools project is to
> provide FreeBSD's developers and system administrators with
> non-intrusive, low-overhead and innovative ways of measuring and
> analysing system performance" your website. Have you ever measured the
> performance impact of such tools?

No, I didn't, I just did the port.

But maybe Joseph Koshy (who is the author of PMCTools) has measurements of
the PMC library on different machines/environments. See or contact him directly.

The port itself is based directly on the PMCtools (i.e. it's almost a
wrapper to convert PAPI calls into PMC calls), so I don't think that PAPI
adds too much overhead to this basic library.

I'm interested to run such tool on production machines in the future but
> only if the performance impact isn't that high.
> Regards,
> Thomas
> Harald Servat wrote:
> > Hello,
> >
> >  I'm glad to announce you that PAPI-3.5.0 has reached the FreeBSD ports
> > tree and now it's generally available for all FreeBSD users.
> >
> >  Port information is available at
> >
> >
> >  See for
> > installation instructions.
> >
> >  There are some issues with P4 processors that need to be fixed on
> > PAPI_write / PAPI_reset routines, but the package have the minimal (and
> > most
> > important functionality) working fine for the rest of the substrates.
> >
> > Regards,

Empty your memory,
with a free()...
like a pointer!

If you cast a pointer to an integer,
it becomes an integer,
if you cast a pointer to a struct,
it becomes a struct.

The pointer can crash...,
and can overflow.

Be a pointer my friend...

More information about the freebsd-performance mailing list