On 28-Aug-2000 Linus Torvalds wrote:
> I think the right solution is to completely split up "schedule()" into two
> different functions (which just share 99% of the code), and basically have
> the idle thread call the _other_ schedule. The one that never does the
> test at all.
>
> That way you know statically whether you are the idle thread or not. No
> test at all.
I thought about it some more. Yes, I can tell statically if I am the idle
thread or not, but how do I tell if the process that was running before me on
the CPU is the idle thread? That's what __schedule_tail() cares about,
prev. If we have a separate schedule for idlers they can avoid the test in
their __schedule_tail but the generic case still has to check. As far as I
can tell having a separate schedule would allow idle threads to skip some of
the early tests in schedule() but I doubt it makes sense from a cache
footprint POV.
Speaking of footprints, that reschedule_idle should really be turned into a
FASTCALL. It's large and inlining creates too many copies in frequently
executed code. __wake_up() even has two of them, check how large it is some
time.
-- Dimitris Michailidis dimitris@engr.sgi.com - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Aug 31 2000 - 21:00:22 EST