2.2.14 feels slow and non-interactive.

[Andrea Arcangeli]

On Fri, 21 Jan 2000, Alan Cox wrote:

>Im not sure 2.2.13->14 is the change, its certainly involved. The drivers/stack
>havent all changed, its not a driver issue. 2.2.13 also certainly shows parts
>of the behaviour.

If it's not a driver issue I don't have other ideas right now. Nothing
relevant to VM performances is changed in recent 2.2.x kernels as far I
can tell.

>If you think the swapping diffs from 2.2.14aa1 will help then let me know
>where to grab just the needed bits and we'll try it for 2.2.15pre4.

Ok, for 2.2.15pre I suggest these two patches:

	ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/kernels/v2.2/2.2.14aa1/no-swapout-2.2.10-B.gz
	ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/kernels/v2.2/2.2.14aa1/trashing-mem-2.2.10-A.gz

The first one will make sure to not swapout/swapin at infinity during
heavy I/O on disk. The second one will avoid an hog to stall the machine.

And I would suggest also this below patch even if way less important and
it may hurt a bit on 3/4mbyte machines (the few 3/4mbyte machines can
fixup the change via sysctl at boot):

	ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/kernels/v2.2/2.2.14aa1/shrink_all_cache-2.2.10-A.gz

BTW, Alan could you tell me about your problems with atomic allocation? I
think I missed your post (I did a fast search but I didn't succeed to find
it yet, you write lots of emails :-).

Andrea


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo en vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/