Re: swap-prefetch: A smart way to make good use of idle resources (was: updatedb)

[grundig]

El Sat, 28 Jul 2007 02:03:19 +0200 (CEST), "Indan Zupancic" <indan@xxxxxx> escribió:

> Perhaps one of the reasons is that this is core kernel code. And that it isn't a new
> feature, but a performance improvement with doubtful trade-offs. The problem
> statement isn't clear either. It seems like a natural enhancement, but is that enough
> reason to merge it? Maybe, maybe not. But if slow swap-in is the problem, shouldn't
> that be fixed instead of bypassed?


The problem AFAIK it's clear: Swap in is slow by nature because it's random. Well, it "looks"
like it should be random. Now that would be a interesting thing to analyze: locality of the
swapped-in stuff for different worloads - if the swapped-in stuff has some locality, then
it certainly should possible to fix the swapin code to do some kind of "swapahead". But if
the stuff is swapped in completely randomly, only stuff like swap prefetch is going to help.

The good thing about swap prefetch is that it fixes the problem today without forbidding
future improvements to the VM.

-
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/