Re: sched_clock() uses are broken

[Nick Piggin]

Mike Galbraith wrote:

> On Mon, 2006-05-08 at 14:46 +1000, Nick Piggin wrote:
>
> > It should either get ripped out, or perhaps converted to use jiffies
> > until a sane high resolution, low overhead scheme is developed (if
> > ever). And that would exclude something that does this accounting in
> > fastpaths for the 99.99% of processes that never use it.
>
> The accounting is really light compared to the interactivity part.  That
> doesn't need to be in the fast path, and in my tree it isn't.

And it is worse than useless if it is wrong.

> FWIW (0), yy tree is missing every last shred of the interactivity code,
> and not missing it one bit.

Join the club ;)

> [root@Homer]:> diffstat xx
> sched.c |  480
> +++++++++++++++++++++++++++++-----------------------------------
> 1 files changed, 219 insertions(+), 261 deletions(-)
>
> And that's with full throttling, and absolute starvation proofing.
>
> Ho hum.  Back to work on my never-going-anywhere-but-fun tree :)

Join the club ;)

Nah, the interactivity stuff isn't too bad (as a function of bug reports
to lkml). One day when there is nothing else wrong with the kernel, it will
probably end up getting reworked. No need to stir it up now though, really.

--

Send instant messages to your online friends http://au.messenger.yahoo.com 
-
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/