Alex Bligh - linux-kernel wrote:
>
> --On Tuesday, September 11, 2001 6:07 PM +0200 Daniel Phillips
> <phillips@bonn-fries.net> wrote:
>
> > There is clearly something nonoptimal about the hardware readahead and/or
> > caching.
>
> Partly I guess that the disk h/w cache is the wrong side of a potential
> IO bottleneck - i.e. you can read faster from main memory than from
> any disk cache.
But during the read from the disk you can do different things in
paralell.
Think DMA! And then please remember my posts from a year ago:
We have already 3 different read-ahead layers in Linux!
Read it twice THRE one on top of the other:
1. Physical.
2. Inside the driver.
2a. block device layer.
3. File system layer.
And in addition to this there is:
2a. or maybe even 2b if you take MD and RAID into account as well.
And plase remember that each of them can fight against the other becouse
they all presume different ordering mechanisms.
-
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/
This archive was generated by hypermail 2b29 : Sat Sep 15 2001 - 21:00:34 EST