> Chris Mason wrote:
>
> >>I don't mean to suggest that the dentry cache locking is an easy problem to solve, but the problem discussed is a real one, and it is sufficient to illustrate that the unified cache is fundamentally flawed as an algorithm compared to using subcache plugins.
> >>
> >
> >It isn't just dentries. If a subcache object is in use, it can't be moved
> >to a warmer page without invalidating all existing pointers to it.
> >
> >If it isn't in use, it can be migrated when the VM asks for the page to
> >be flushed.
> >
> garbage collection is a lot of work to implement --- there are a lot of
> good reasons why ext2 doesn't shrink directories.....;-)
>
> really guys, you can get me to agree that it is more work to code, you
> can even get me to agree to skip it for now because we are all busy, but
> the design principle remains valid --- using per page aging of subpage
> objects that do not correlate in accesses leads to diffused hot sets,
> and that means that the cache will perform as though it was much smaller
> than it is.
Can we get you to agree that basically all subpage objects are immovable?
And as a consequence that garbage collecting at subpage levels doesn't
guarantee freeing up any pages that can then be given up to other
subsystems in response to VM pressure? The GC must think in terms of pages
to actually make progress.
One of the design goals of slab by the way is that objects of a similar
type will end up having similar lifetimes, avoiding some of the worst
cases of sub-page allocations.
-- "Love the dolphins," she advised him. "Write by W.A.S.T.E.."- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/