Re: [Announce] [patch] Modular Scheduler Core and Completely Fair Scheduler [CFS]

[William Lee Irwin III <wli@holomorphy.com>]

On Mon, Apr 16, 2007 at 11:50:03PM -0700, Davide Libenzi wrote:
> I had a quick look at Ingo's code yesterday. Ingo is always smart to 
> prepare a main dish (feature) with a nice sider (code cleanup) to Linus ;)
> And even this code does that pretty nicely. The deadline designs looks 
> good, although I think the final "key" calculation code will end up quite 
> different from what it looks now.

The additive nice_offset breaks nice levels. A multiplicative priority
weighting of a different, nonnegative metric of cpu utilization from
what's now used is required for nice levels to work. I've been trying
to point this out politely by strongly suggesting testing whether nice
levels work.


On Mon, Apr 16, 2007 at 11:50:03PM -0700, Davide Libenzi wrote:
> I would suggest to thoroughly test all your alternatives before deciding. 
> Some code and design may look very good and small at the beginning, but 
> when you start patching it to cover all the dark spots, you effectively 
> end up with another thing (in both design and code footprint).
> About O(1), I never thought it was a must (besides a good marketing 
> material), and O(log(N)) *may* be just fine (to be verified, of course).

The trouble with thorough testing right now is that no one agrees on
what the tests should be and a number of the testcases are not in great
shape. An agreed-upon set of testcases for basic correctness should be
devised and the implementations of those testcases need to be
maintainable code and the tests set up for automated testing and
changing their parameters without recompiling via command-line options.

Once there's a standard regression test suite for correctness, one
needs to be devised for performance, including interactive performance.
The primary difficulty I see along these lines is finding a way to
automate tests of graphics and input device response performance. Others,
like how deterministically priorities are respected over progressively
smaller time intervals and noninteractive workload performance are
nowhere near as difficult to arrange and in many cases already exist.
Just reuse SDET, AIM7/AIM9, OAST, contest, interbench, et al.


-- wli
-
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/