On Fri, Sep 28, 2012 at 05:23:27PM +0200, Peter Zijlstra wrote:
> On Fri, 2012-09-28 at 16:53 +0200, Andi Kleen wrote:
> >
> > Also as a selfish reason I would prefer something that is short
> > to type. The qualifiers are quite common in scripts that do
> > measurements here. So I would prefer t
On Fri, 2012-09-28 at 17:29 +0200, Andi Kleen wrote:
> Still would need new fields, but I presume that could be fit in.
>
Very much so, for all except the P4 PMU the first config is directly
mapped to the perfctl msr.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
th
On Fri, Sep 28, 2012 at 05:19:19PM +0200, Peter Zijlstra wrote:
> On Fri, 2012-09-28 at 16:53 +0200, Andi Kleen wrote:
> > If I moved this into sysfs this would imply that the perf stat -T
> > code would become Haswell specific. As far as I understand normally
> > you guys don't want things like t
On Fri, 2012-09-28 at 16:53 +0200, Andi Kleen wrote:
>
> Also as a selfish reason I would prefer something that is short
> to type. The qualifiers are quite common in scripts that do
> measurements here. So I would prefer to keep :t and :c as user
> interface. But the internal implementation can
On Fri, 2012-09-28 at 16:53 +0200, Andi Kleen wrote:
> If I moved this into sysfs this would imply that the perf stat -T
> code would become Haswell specific. As far as I understand normally
> you guys don't want things like that. Would everyone be ok with
> having specific code there?
Have a lo
On Fri, Sep 28, 2012 at 11:02:00AM +0200, Peter Zijlstra wrote:
> On Thu, 2012-09-27 at 21:31 -0700, Andi Kleen wrote:
> > + intx : 1, /* count inside
> > transaction */
> > + intx_checkpointed : 1, /* checkpointed in
> > tra
On Fri, Sep 28, 2012 at 11:02 AM, Peter Zijlstra wrote:
> On Thu, 2012-09-27 at 21:31 -0700, Andi Kleen wrote:
>> + intx : 1, /* count inside
>> transaction */
>> + intx_checkpointed : 1, /* checkpointed in
>> transaction */
On Thu, 2012-09-27 at 21:31 -0700, Andi Kleen wrote:
> + intx : 1, /* count inside
> transaction */
> + intx_checkpointed : 1, /* checkpointed in
> transaction */
I really hate those names.. what are they called in transact
From: Andi Kleen
Expose INTX (count in transaction only, :t) and INTX_CHECKPOINTED
(on transaction abort restore counter, :c) attributes as generic perf event
attributes. These are important for measuring basic hardware transactional
behaviour.
They also need to be handled in a special way in t
9 matches
Mail list logo
