2017-11-27 16:37 GMT+08:00 Michal Hocko <mhocko@xxxxxxxx>:
> On Mon 27-11-17 16:32:42, Yafang Shao wrote:
>> 2017-11-27 16:29 GMT+08:00 Yafang Shao <laoar.shao@xxxxxxxxx>:
>> > 2017-11-27 16:21 GMT+08:00 Michal Hocko <mhocko@xxxxxxxx>:
>> >> On Mon 27-11-17 16:06:50, Yafang Shao wrote:
>> >>> +cc fcicq
>> >> [...]
>> >>> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
>> >>> index 8a15511..6c5c018 100644
>> >>> --- a/mm/page-writeback.c
>> >>> +++ b/mm/page-writeback.c
>> >>> @@ -377,7 +377,16 @@ static unsigned long global_dirtyable_memory(void)
>> >>> if (!vm_highmem_is_dirtyable)
>> >>> x -= highmem_dirtyable_memory(x);
>> >>>
>> >>> - return x + 1; /* Ensure that we never return 0 */
>> >>> + /*
>> >>> + * - Why 100 ?
>> >>> + * - Because the return value will be used by dirty ratio and
>> >>> + * dirty background ratio to calculate dirty thresh and bg thresh,
>> >>> + * so if the return value is two small, the thresh value maybe
>> >>> + * calculated to 0.
>> >>> + * As the max value of ratio is 100, so the return value is added
>> >>> + * by 100 here.
>> >>> + */
>> >>> + return x + 100;
>> >>
>> >> No. We should just revert 0f6d24f87856 ("mm/page-writeback.c: print a
>> >> warning if the vm dirtiness settings are illogical") because it is of a
>> >> dubious value and it causes problems. I am not even sure why it got
>> >> merged. It doesn't have any ack or review and I remember objecting to
>> >> the patch previously as pointless.
>> >> --
>> >
>> > It is reviewed and merged by Andrew.
>> >
>> > From Andrew:
>> >> I think this means that a script which alters both dirty_bytes and
>> > dirty_background_bytes must alter dirty_background_bytes first if they
>> > are being decreased and must alter dirty_bytes first if they are being
>> > increased. Or something like that.
>> >
>> > It will help us to find the error if we don't change these values like this.
>> >
>>
>> And actually it help us find another issue that when availble_memroy
>> is too small, the thresh and bg_thresh will be 0, that's absolutely
>> wrong.
>
> Why is it wrong?
> --
For example, the writeback threads will be wakeup on every write.
I don't think it is meaningful to wakeup the writeback thread when the
dirty pages is very low.
Thanks
Yafang
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
