Messages in this thread | | | Date | Thu, 12 Sep 2002 08:27:29 +0200 | From | Vojtech Pavlik <> | Subject | Re: [PATCH] USER_HZ & NTP problems |
| |
On Wed, Sep 11, 2002 at 03:56:54PM -0700, george anzinger wrote: > Horst von Brand wrote: > > > > Rolf Fokkens <fokkensr@fokkensr.vertis.nl> said: > > > On Tuesday 10 September 2002 08:09, Vojtech Pavlik wrote: > > > > Actually, the clock true frequency is 1193181.8 Hz, although most > > > > manuals say 1.19318 MHz, which, because truncating more digits, also > > > > correct. But 1193180 Hz isn't. If you're trying to count the time > > > > correctly, you should better use 1193182 Hz if staying in integers. > > > > > > I copied the clock frequency from the kernel source, timex.h defines: > > > > > > #define CLOCK_TICK_RATE 1193180 > > > > > > If what you're saying is correct, timex.h uses the wrong value as wel I guess. > > > > I'd be _really_ surprised if that number is within 1% (+-11932 or so) > > correct, an error in the last digit (0.0001%) is surely lost. > > -- > But it DOES explain why the ACPI pm timer at 3579545 is not > exactly 3 * CLOCK_TICK_RATE. Inquiring minds want to know > :)
Yeah. It's all based on NTSC dotclock from the old IBM PC. (which is 4x ACPI clock and 12x PIT clock)
-- Vojtech Pavlik SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |