[PATCH v3] oom, oom_reaper: do not enqueue same task twice

[Tetsuo Handa]

On 2019/01/27 20:40, Michal Hocko wrote:
> On Sun 27-01-19 19:56:06, Tetsuo Handa wrote:
>> On 2019/01/27 17:37, Michal Hocko wrote:
>>> Thanks for the analysis and the patch. This should work, I believe but
>>> I am not really thrilled to overload the meaning of the MMF_UNSTABLE.
>>> The flag is meant to signal accessing address space is not stable and it
>>> is not aimed to synchronize oom reaper with the oom path.
>>>
>>> Can we make use mark_oom_victim directly? I didn't get to think that
>>> through right now so I might be missing something but this should
>>> prevent repeating queueing as well.
>>
>> Yes, TIF_MEMDIE would work. But you are planning to remove TIF_MEMDIE. Also,
>> TIF_MEMDIE can't avoid enqueuing many threads sharing mm_struct to the OOM
>> reaper. There is no need to enqueue many threads sharing mm_struct because
>> the OOM reaper acts on mm_struct rather than task_struct. Thus, enqueuing
>> based on per mm_struct flag sounds better, but MMF_OOM_VICTIM cannot be
>> set from wake_oom_reaper(victim) because victim's mm might be already inside
>> exit_mmap() when wake_oom_reaper(victim) is called after task_unlock(victim).
>>
>> We could reintroduce MMF_OOM_KILLED in commit 855b018325737f76
>> ("oom, oom_reaper: disable oom_reaper for oom_kill_allocating_task")
>> if you don't like overloading the meaning of the MMF_UNSTABLE. But since
>> MMF_UNSTABLE is available in Linux 4.9+ kernels (which covers all LTS stable
>> versions with the OOM reaper support), we can temporarily use MMF_UNSTABLE
>> for ease of backporting.
> 
> I agree that a per-mm state is more optimal but I would rather fix the
> issue in a clear way first and only then think about an optimization on
> top. Queueing based on mark_oom_victim (whatever that uses to guarantee
> the victim is marked atomically and only once) makes sense from the
> conceptual point of view and it makes a lot of sense to start from
> there. MMF_UNSTABLE has a completely different purpose. So unless you
> see a correctness issue with that then I would rather go that way.
> 

Then, adding a per mm_struct flag is better. I don't see the difference
between reusing MMF_UNSTABLE as a flag for whether wake_oom_reaper() for
that victim's memory was already called (what you think as an overload)
and reusing TIF_MEMDIE as a flag for whether wake_oom_reaper() for that
victim thread can be called (what I think as an overload). We want to
remove TIF_MEMDIE, and we can actually remove TIF_MEMDIE if you stop
whack-a-mole "can you observe it in real workload/program?" game.
I don't see a correctness issue with TIF_MEMDIE but I don't want to go
TIF_MEMDIE way.