Re: [PATCH V2 1/6] perf,core: allow invalid context events to be part of sw/hw groups

From: Mark Rutland
Date: Fri Apr 17 2015 - 05:48:00 EST


On Thu, Apr 16, 2015 at 10:23:42PM +0100, Andi Kleen wrote:
> > From my PoV that makes sense. One is CPU-affine, the other is not, and
> > the two cannot be scheduled in the same PMU transaction by the nature of
> > the hardware. Fundamentally, you cannot provide group semantics due to
> > this.
>
> Actually you can. Just use it like a free running counter, and the
> different groups sample it. This will work from the different CPUs,
> as long as the event is the same everywhere.

... which would give you arbitrary skew, because one counter is
free-running and the other is not (when scheduling a context in or out we stop
the PMU)

>From my PoV that violates group semantics, because now the events aren't
always counting at the same time (which would be the reason I grouped
them in the first place).

> The implemention may not be quite right yet, but the basic concept
> should work, and is useful.

I can see that associating counts from different PMUs at points in time
may be useful, even if they aren't sampled at the precise same time, and
you have weaker guarantees than the current group semantics.

However, it is the case that you cannot offer group semantics.

Mark.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/