Re: [PATCH 0/4] exit: Make unlikely case in mm_update_next_owner() more scalable

[Eric W. Biederman]

Kirill Tkhai <ktkhai@xxxxxxxxxxxxx> writes:

> On 01.06.2018 18:25, Eric W. Biederman wrote:
>> Michal Hocko <mhocko@xxxxxxxxxx> writes:
>> 
>>> On Fri 01-06-18 09:32:42, Eric W. Biederman wrote:
>>>> Michal Hocko <mhocko@xxxxxxxxxx> writes:
>>> [...]
>>>>> Group leader exiting early without tearing down the whole thread
>>>>> group should be quite rare as well. No question that somebody might do
>>>>> that on purpose though...
>>>>
>>>> The group leader exiting early is a completely legitimate and reasonable
>>>> thing to do, even if it is rare.
>>>
>>> I am not saying it isn't legitimate. But the most common case is the
>>> main thread waiting for its threads or calling exit which would tear the
>>> whole group down. Is there any easy way to achieve this other than tkill
>>> to group leader? Calling exit(3) from the leader performs group exit
>>> IIRC.
>> 
>> pthread_exit from the group leader.
>> 
>>> I am not arguing this is non-issue. And it certainly is a problem once
>>> somebody wants to be nasty... I was more interested how often this
>>> really happens for sane workloads.
>> 
>> That is a fair question.  All I know for certain is that whatever Kirill
>> Tkhai's workload was it was triggering this the slow path.
>
> It was triggered on a server, where many VPS of many people are hosted.
> Sorry, I have no an idea what they did.

That at least tells us it was naturally occurring.  Which makes this a
real problem in the real world.

Eric