From 81135899313d3c76f92e08463d2070e0a08eb817 Mon Sep 17 00:00:00 2001 From: Bob Dougherty Date: Wed, 25 Oct 2000 15:31:51 +0000 Subject: [PATCH] (bob dougherty) Fixed typos in temp-conversion polynomial fit formula. (thanks to Uwe Anderson for finding them.) git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@914 7894878c-1315-0410-8ee3-d5d059ff63e0 --- kernel/chips/via686a.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/kernel/chips/via686a.c b/kernel/chips/via686a.c index 6d513f6d..dc4e26a0 100644 --- a/kernel/chips/via686a.c +++ b/kernel/chips/via686a.c @@ -199,8 +199,11 @@ extern inline u8 FAN_TO_REG(long rpm, int div) // Kaam ) a bit better. It also give more reasonable // numbers on my machine (ie. they agree with what my BIOS tells me). // Here's the fifth-order fit to the 8-bit data: -// temp = 1.625093e-09*val^5 - 1.001632e-06*val^4 + 2.457653e-04*val^3 - -// 2.967619e-02*val^2 + 2.175144e+00*val - 7.090067e+0. +// temp = 1.625093e-10*val^5 - 1.001632e-07*val^4 + 2.457653e-05*val^3 - +// 2.967619e-03*val^2 + 2.175144e-01*val - 7.090067e+0. +// +// (2000-10-25- RFD: thanks to Uwe Andersen for +// finding my typos in this formula!) // // Alas, none of the elegant function-fit solutions will work because we // aren't allowed to use floating point in the kernel and doing it with