Messages in this thread | | | From | Giuseppe Bilotta <> | Subject | Re: [PATCH] mm: implement swap prefetching (v26) | Date | Wed, 1 Mar 2006 22:34:01 +0100 |
| |
[Hope I managed to keep the CC list intact ... I'm using a newsreader to follow LKML from gmane]
On Sat, 18 Feb 2006 01:26:57 +1100, Con Kolivas wrote:
> On Saturday 18 February 2006 01:18, Kevin Radloff wrote: >> On 2/17/06, Con Kolivas <kernel@kolivas.org> wrote: >>> Added disabling of swap prefetching when laptop_mode is enabled. >> >> Why bother with this? As someone commented in a previous thread, >> wouldn't it be better to let the laptop_mode script handle it? > > The discussion was about what size to make the swap prefetching. Since the > size is not user tunable any more that is not the case. I had an offlist > discussion with Bart Samwel about it and basically if your drive spins down > at 5 seconds (which is what commonly happens with laptop mode) you will never > have an opportunity to prefetch. This means swap prefetch will basically > always spin up the drive nullifying laptop mode. On balance if you care about > power more than anything to actually set laptop mode I suspect you wont want > prefetch using any more power.
Would it make any sense to just delay swap prefetch execution and/or analysis to "as soon as the HD is spun up/just before the HD is spun down", rather than completely disabling it?
-- Giuseppe "Oblomov" Bilotta
Hic manebimus optime
- 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/
| |