Re: [PATCH]: 1/4 batch mark_page_accessed()

[Marcelo Tosatti]

On Sat, Nov 27, 2004 at 01:41:58PM +0300, Nikita Danilov wrote:
> Marcelo Tosatti writes:
> 
> [...]
> 
>  > 
>  > Average Half Load -j 4 Run:                     Average Half Load -j 4 Run:
>  > Elapsed Time 274.916                            Elapsed Time 245.026
>  > User Time 833.63                                User Time 832.34
>  > System Time 73.704                              System Time 73.41
>  > Percent CPU 335.8                               Percent CPU 373.6
>  > Context Switches 12984.8                        Context Switches 13427.4
>  > Sleeps 21459.2                                  Sleeps 21642
> 
> What is "System Time" here? Does in include CPU consumption by, say,
> kswapd?

I dont think "System Time" counts for kswapd - but only time spent inside 
the kernel, including page reclamation efforts.

> If not, then one may conjecture that mark_page_accessed batching
> modifies ordering of pages on zone "LRU" lists in a way that measurably
> changes efficiency of VM scanning, and mostly affects process that scans
> most---kswapd.

I agree.
-
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/