Re: [PATCH]: perf/core: addressing 4x slowdown during per-process profiling of STREAM benchmark on Intel Xeon Phi

[Peter Zijlstra]

On Mon, May 29, 2017 at 05:22:54PM +0200, Peter Zijlstra wrote:
> On Mon, May 29, 2017 at 04:43:09PM +0300, Alexey Budankov wrote:
> > On 29.05.2017 15:03, Alexander Shishkin wrote:
> > > Alexey Budankov <alexey.budankov@xxxxxxxxxxxxxxx> writes:
> 
> > > > +		} else if (event->cpu > node_event->cpu) {
> > > > +			node = &((*node)->rb_right);
> > > > +		} else {
> > > > +			list_add_tail(&event->group_list_entry,
> > > > +					&node_event->group_list);
> > > 
> > > So why is this better than simply having per-cpu event lists plus one
> > > for per-thread events?
> > 
> > Good question. Choice of data structure and layout depends on the operations
> > applied to the data so keeping groups as a tree simplifies and improves the
> > implementation in terms of scalability and performance. Please ask more if
> > any.
> 
> Since these lists are per context, and each task can have a context,
> you'd end up with per-task-per-cpu memory, which is something we'd like
> to avoid (some archs have very limited per-cpu memory space etc..).
> 
> Also, we'd like to have that tree for other reasons, like for instance
> that heterogeneous PMU crud ARM has. Also, with a tree we can easier do
> time based round-robin scheduling,
> 

Oh and in general multi-PMU stuff, aside from hetero PMU becomes much
easier.