Re: [PATCH] mm: moving dirty pages balancing to pdfludh entirely

[Ananiev, Leonid I]

 Nikita Danilov writes:
> Wouldn't this interfere with current->backing_dev_info logic?

The proposed patch does not modify that logic.

> Maybe pdflush threads should set this field too?

If pdflush sets current->backing_dev_info it means to ignore congested
device state.
It will not help to get reclaimed memory for pdflush but will increase
congestion, and pdflush will be dependent from swapping out process
using congested queue. The queue may be just congested because of memory
for queue. Other page could be better candidate for memory reclamation.

So, pdflush should not set this field.

Leonid
-
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/