Dear Luca
Thank you for the offer.
The issue is that I still spend too much time on that issue and
I don't want to blow hours planned for that task.
Therefore, I will disable monitoring of the fan.
It seems that Asus boards do not suit professional needs
if such essential features do not work correctly.
Best regards
Sascha
Zitat von Luca Tettamanti <kronos.it@xxxxxxxxx>:
On Wed, Sep 18, 2013 at 12:28 PM, Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
On Wed, Sep 18, 2013 at 10:51:37AM +0200, Sascha Haupt wrote:
Dear @ll,
I am monitoring multiple systems (with Asus M5A78LE mainboards)
using sensors and Munin.
On one node I often get a value of 675000 RPM for the FAN2 speed.
Usually, the value of FAN2 is approx 1950 RPM.
The jump can frequently be seen if sensors is executed manually in a
fast sequence:
The driver only reports what it gets from the system through ACPI.
I don't think there is anything we can do about it.
Luca, any idea ?
Usually the ACPI code tries to read the speed from the sensor a few
times until it "makes sense" (according to some board specific logic);
if it fails it returns garbage.
If you send me the DSDT (/sys/firmware/acpi/tables/DSDT) I'll take a look.
In theory you could rewrite the method and replace the DSDT, but I
don't think it's worth the effort.
Luca
_______________________________________________
lm-sensors mailing list
lm-sensors@xxxxxxxxxxxxxx
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
