[Intel-gfx] [PATCH] [RFC] mm/shrinker: Add a shrinker flag to always shrink a bit

[Knut Petersen]

> Looking at the patch which introduced these error message for you, which
> changed the ->count_objects return value from 0 to SHRINK_STOP your patch
> below to treat 0 and SHRINK_STOP equally simply reverts the functional
> change.

Yes, for i915* it de facto restores the old behaviour.

> I don't think that's the intention behind SHRINK_STOP. But if it's the
> right think to do we better revert the offending commit directly.
But there is other code that also returns SHRINK_STOP. So i believe it´s better to
adapt shrink_slab_node() to handle SHRINK_STOP properly than to revert 81e49f.

>   And since I lack clue I think that's a call for core mm guys to make.

I agree. They´ll probably have to apply some additional changes to
shrink_slab_node(). It really doesn´t look right to me, but they certainly
know better what the code is supposed to do ;-)


cu,
  Knut