Re: [PATCH] x86, perf: Use INST_RETIRED.PREC_DIST for cycles:pp on Skylake

[Andi Kleen]

> Right, but suppose one user (root) would be sampling machine wide with
> PEBS whilst another user is using it just on his one task.

The root case is likely perf top which does not use PEBS.

> > > And you having killed the flag bits for PEBS events precludes people
> > > from using this manually, right?  I think we want to exempt .inv=1
> > > .cmask=16 from that general rule on general utility value.
> > 
> > I don't want to do this for every event. It has caused problems
> > in the past. There's also no reason to use the other events.
> 
> But it does allow users to play around with alternative encodings for
> cycles etc.. We've made use of this in the past when SNB turned out to
> have broken the event.

If they want to violate the SDM they can hack the kernel.
It's risky to do as you well know.

> > > We could maybe abuse .precise_ip = 3 for this?
> > 
> > I implemented this now, even though it's ugly.
> 
> Agreed, its not ideal.

So it turns out that UOPS_RETIRED.ALL+pebs only works by accident
on Skylake. It's not in the specification.

So on Skylake it has to be unconditional for any cycles:p
On earlier parts it can be only for three p to handle your 
unlikely scenario.

-Andi

-- 
ak@xxxxxxxxxxxxxxx -- Speaking for myself only
--
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/