|| ||Alan Cox <alan-AT-lxorguk.ukuu.org.uk>|
|| ||Linus Torvalds <torvalds-AT-linux-foundation.org>|
|| ||Re: [PATCH 4/5] select: make select() use schedule_hrtimeout()|
|| ||Fri, 29 Aug 2008 17:11:08 +0100|
|| ||Arjan van de Ven <arjan-AT-infradead.org>,
linux-kernel-AT-vger.kernel.org, mingo-AT-elte.hu, tglx-AT-tglx.de|
> "schedule_timeout()", there's a big difference between asking for two
> ticks and asking for two seconds. The latter should probably try to round
> to a nice timer tick basis for power reasons).
I disagree - that is fixing the problem in the wrong place. The timer
structure needs an accuracy field of some form that the existing timer
functions initialise to 0.
On a heavily loaded system with things like network events the behaviour
of the overall system is too complex to do the job well except at the
timer level which seems all. At the virtualisation level the hypervisor
needs to be doing the work to merge timer events between guests for power
Once the timers have an accuracy representation there doesn't need to be
any real difference anywhere in the stack.
to post comments)