Re: [PATCH] mm,oom: fix oom invocation issues

[Michal Hocko]

On Thu 18-05-17 16:29:01, Michal Hocko wrote:
> On Thu 18-05-17 22:57:10, Tetsuo Handa wrote:
[...]
> > Anyway, I want
> > 
> > 	/* Avoid allocations with no watermarks from looping endlessly */
> > -	if (test_thread_flag(TIF_MEMDIE))
> > +	if (alloc_flags == ALLOC_NO_WATERMARKS && test_thread_flag(TIF_MEMDIE))
> > 		goto nopage;
> > 
> > so that we won't see similar backtraces and memory information from both
> > out_of_memory() and warn_alloc().
> 
> I do not think this is an improvement and it is unrelated to the
> discussion here.

I am sorry, I've misread the diff. It was the comment below the diff
which confused me. Now that I looked at it again it actually makes sense.
I would still like to get rid of out_of_memory from pagefault_out_of_memory
but doing the above sounds safer for the stable backport. Care to create
a proper patch with the full changelog, please?
-- 
Michal Hocko
SUSE Labs