> We could implement some kind of priority sorted waitqueue.
> I read about some prio aware waitqueue patch (from Ingo Molnar) used
> by a NGPT developer for the futexes (I.Gonzalez?)
I am just a contributor to NGPT ... what I did was to apply the
concept in Ingo's O(1) scheduling to the selection of tasks waiting
for the futex event.
> I thought about some waitqueue in the kernel, that at least put
> realtime processes at the front of the waitqueue.
>
> It would be nice to have some generic mechanism in the kernel since
> only prio aware message read feature does not buy you much, IMHO
I did something like that also for the futexes [priority based futexes].
Check the attached patch vs. 2.5.45 [sorry for not inlining it, but I
am w/ Outlook now and it breaks it].
The nice next step would be to go ahead and create a breed of wait queues
that supported priority-based wake up, but there seemed to be not too
much interest; in any case, we'd need to come out with a sollution that
would not waste 2*sizeof(void*) * NUM_PRIO bytes per wait-queue, but
still being O(1) [and now that we are at asking, if somebody knows where
to buy a brand new Mercedes Benz C320 at 25%, pls email me]
Inaky Perez-Gonzalez -- Not speaking for Intel - opinions are my own [or my
fault]
This archive was generated by hypermail 2b29 : Sat Nov 23 2002 - 22:00:28 EST