Re: [patch -mmotm] mm: invoke oom killer for __GFP_NOFAIL

[Minchan Kim]

On Tue, May 12, 2009 at 4:32 AM, David Rientjes <rientjes@xxxxxxxxxx> wrote:
> On Mon, 11 May 2009, Minchan Kim wrote:
>
>> But this patch increases the probability of innocent task killing.
>> Is GFP_NOFAIL's early bailout more important than killing of innocent task ?
>>
>
> __GFP_NOFAIL's bailout is, by definition, a bug in the page allocator and
> this patch fixes it.
>

What I mean is If fortunately other process have a lot of memory is
exited while looping, __GFP_NOFAIL can return.
I have no objection about your patch.
Just, I have a concern of killing innocent task.

As Mel and Andrew pointed out, fundamentally, __GFOP_NOFAIL and
high-order allocation itself is bad code.
Okay. we have to insert WARN_ON properly.

Thanks for good contribution. :)

-- 
Kinds regards,
Minchan Kim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/