Re: [PATCH v2]: perf/core: addressing 4x slowdown during per-process, profiling of STREAM benchmark on Intel Xeon Phi

[Alexey Budankov <alexey.budankov@linux.intel.com>]

Hi Peter,

On 21.06.2017 18:39, Alexey Budankov wrote:
> 
> Hi,
> 
> On 15.06.2017 20:42, Alexey Budankov wrote:
>> On 29.05.2017 14:45, Alexey Budankov wrote:
>>> On 29.05.2017 14:23, Peter Zijlstra wrote:
>>>> On Mon, May 29, 2017 at 01:56:05PM +0300, Alexey Budankov wrote:
>>>>> On 29.05.2017 13:43, Peter Zijlstra wrote:
>>>>
>>>>>> Why can't the tree do both?
>>>>>>
>>>>>
>>>>> Well, indeed, the tree provides such capability too. However switching to
>>>>> the full tree iteration in cases where we now go through _groups lists will
>>>>> enlarge the patch, what is probably is not a big deal. Do you think it is
>>>>> worth implementing the switch?
>>>>
>>>> Do it as a series of patches, where patch 1 introduces the tree, patches
>>>> 2 through n convert the list users into tree users, and patch n+1
>>>> removes the list.
>>>
>>> Well ok, let's do that additionally but please expect delay in delivery (I am OOO till Jun 14).
>>
>> addressed in v3.
>>
>>>
>>>>
>>>> I think its good to not have duplicate data structures if we can avoid
>>>> it.
>>>>
>>>
>>> yeah, makes sense.
>>>
>>>
>>>
>>
>>
> 
> After straightforward switch from struct list_head to struct rb_tree for flexible_groups I now get dmesg dumps on rb tree corruptions. That happens when iterating thru tree instead of thru list. No additional
> synchronization for the tree access was added. It looks like there are
> some assumptions on the list_head type in the implementation itself.
> 
> Are there any ideas on why that corruptions may happen?
> 
> I still suggest isolating event groups into a separate object (please see patch v4-1/4):
> 
> struct perf_event_groups {
>     struct rb_root     tree;
>     struct list_head list;
> };
> 
> struct perf_event_context {
> ...
> struct perf_event_groups pinned_groups;
> struct perf_event_groups flexible_groups;
> 
> and implementing new API for the object:
> 
> perf_event_groups_empty()
> perf_event_groups_init()
> perf_event_groups_insert()
> perf_event_groups_delete()
> perf_event_groups_rotate(..., int cpu)
> perf_event_groups_iterate_cpu(..., int cpu)
> perf_event_groups_iterate()
> 
> so that perf_event_groups_iterate() would go thru list but leaving
> the opportunity of iteration thru tree for a separate patch because
> complete transition to rb trees may incur synchronization overhead in runtime.

Completely got rid of list and tree duplication in patch v5 4/4.
Please see here: 

[PATCH v5 4/4] perf/core: addressing 4x slowdown during per-process 
                          profiling of STREAM benchmark on Intel Xeon Phi

> 
> Thanks,
> Alexey
> 

Thanks,
Alexey