On Thu 15-09-16 10:41:18, Johannes Weiner wrote:

Hi Michal,

On Thu, Sep 01, 2016 at 11:51:00AM +0200, Michal Hocko wrote:

quoted

Hi,
this is an early RFC to see whether the approach I've taken is acceptable.
The series is on top of the current mmotm tree (2016-08-31-16-06). I didn't
get to test it so it might be completely broken.

The primary point of this series is to get rid of TIF_MEMDIE finally.
Recent changes in the oom proper allows for that finally, I believe. Now
that all the oom victims are reapable we are no longer depending on
ALLOC_NO_WATERMARKS because the memory held by the victim is reclaimed
asynchronously. A partial access to memory reserves should be sufficient
just to guarantee that the oom victim is not starved due to other
memory consumers. This also means that we do not have to pretend to be
conservative and give access to memory reserves only to one thread from
the process at the time. This is patch 1.

Patch 2 is a simple cleanup which turns TIF_MEMDIE users to tsk_is_oom_victim
which is process rather than thread centric. None of those callers really
requires to be thread aware AFAICS.

The tricky part then is exit_oom_victim vs. oom_killer_disable because
TIF_MEMDIE acted as a token there so we had a way to count threads from
the process. It didn't work 100% reliably and had it own issues but we
have to replace it with something which doesn't rely on counting threads
but rather find a moment when all threads have reached steady state in
do_exit. This is what patch 3 does and I would really appreciate if Oleg
could double check my thinking there. I am also CCing Al on that one
because I am moving exit_io_context up in do_exit right before exit_notify.

You're explaining the mechanical thing you are doing, but I'm having
trouble understanding why you want to get rid of TIF_MEMDIE. For one,
it's more code. And apparently, it's also more complicated than what
we have right now.

Can you please explain in the cover letter what's broken/undesirable?

Sure, I will extend the cover when submitting the series again. This RFC
was mostly aimed at correctness so I focused more on technical details.
Patch 1 should contain some reasoning. Do you find it sufficient or I
should extend on top of that?

Thanks!

-- 
Michal Hocko
SUSE Labs

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>