[LTP] [PATCH] memcg_stress: survive OOM by targeting the stressors

[Li Wang]

Andrea Cervesato wrote:

> > What do you think about doing this in the shell lib so all shell
> > tests benefit, rather than only memcg_stress?
> 
> this patch was sent with a wrong assumption indeed. As you can see
> from the log, the OOM message is wrong and I pasted the wrong example :)

Ah, the memcg_stress test already done and kirk start a new
one memcg_control, that oom trigged by there.

> But this raised a good point that I think you spotted quite well. It
> would be nice to implement this indeed, since shell tests are often
> causing issues when it comes to containers or stressing the memory.

Sure, if the shell part encounters same situation, appropriate
measures should be taken.

> I will leave to @Petr the last word because he's more expert than me
> on the shell tests.

+1

-- 
Regards,
Li Wang