Hello Gordon,
Once the real-time core module is loaded, it takes control of the system
timer and __assumes__ Linux doesn't play around with it anymore. The
rt core will __try to__ call upon the normal Linux system timer every
time 10ms have passed (or every 1/HZ, to be accurate), but it may call
it 12ms afterwards. This usually isn't very problematic, but it can be
if you're trying to use the do_gettimeofday() and its sister system call
gettimeofday() as the timings reported may be inaccurate. A while back
I posted a fix for this problem on this list which forces the Linux timer
interrupt to use the TSC available on Pentium-class machines to update
the time, rather than assume that it was last called upon 10ms ago.
Karim
Gordon McNutt wrote:
>
> Hi,
>
> On an x86 (for example), if an RT task programs a clock timer does it
> conflict with the regular system timer interrupts? They both appear to
> use IRQ 0.
>
> --Gordon
>
> -- [rtl] ---
> To unsubscribe:
> echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
> echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
> ---
> For more information on Real-Time Linux see:
> http://www.rtlinux.org/rtlinux/
--
===================================================
Karim Yaghmour
[EMAIL PROTECTED]
Operating System Consultant
(Linux kernel, real-time and distributed systems)
===================================================
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
---
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
