>Without firm timers the wakeups occur on the 30ms and 40ms boundaries
>(nominally 33.3ms). With firm timers, they occur at 33.3ms *every* time
>(excepting long kernel paths).
>
>Something like this really *really* needs to get into the kernel, because
>it's useful for everything from soft-RT to
On Wed, 10 Jul 2002, Kai Vehmanen wrote:
> Not a problem. Timing sensitive user-space apps benefit from this change
> because with 1kHz ticks, kernel will be making scheduling decisions much
> more often.
The solution to a lot of these scheduling problems (non-RT) is firm
timers:
http://www.cse
On Tue, 9 Jul 2002, Paul Sladen wrote:
>>> RTC is still more accurate, but on the other hand, you don't need root
>>> privileges to take advantage of the 1kHz ticks!
> Kernel space only; userspace is still at 100Hz on x86
Not a problem. Timing sensitive user-space apps benefit from this change
On Mon, 8 Jul 2002, Vincent Touquet wrote:
> >
> > RTC is still more accurate, but on the other hand, you don't need root
> > privileges to take advantage of the 1kHz ticks!
>
> can't wait till 2.6 hits the streets :)
Kernel space only; userspace is still at 100Hz on x86
-Paul
--
Nott
Vincent Touquet wrote:
>
> Yum,
>
> can't wait till 2.6 hits the streets :)
breathe deeply
Yum,
can't wait till 2.6 hits the streets :)
vincent
>RTC is still more accurate, but on the other hand, you don't need root
>privileges to take advantage of the 1kHz ticks!
>
>--
> http://www.eca.cx
> Audio software for Linux!
>
>
Here's one interesting commit in the 2.5-tree:
--cut--
D 1.2 02/07/01 14:41:36-07:00 [EMAIL PROTECTED] 3 2 6/4/00020
P include/asm-i386/param.h
C Make Linux/x86 tick at 1kHz instead of 100Hz.
C
C Export the deprecated clock_t frequency (100HZ) as USER_HZ.
--
RTC is still more accurate,
