On Monday 09 June 2014 18:23:31 Pali Rohár wrote: > On Saturday 17 May 2014 22:02:12 Pali Rohár wrote: > > Hello, > > > > I see very annoying bug in sensord. When some hwmon driver > > returns -EINVAL (for any reason), sensord daemon will stop > > reading next sensors values. This means if first hwmon > > device fail to return value, sensord daemon will not log > > anything. > > > > Returning -EINVAL from kernel driver is normal (e.g. when > > graphic card with has internal temperature sensor is turned > > off) and sensord should not fail on this. > > > > Program sensors already handling this problem and will print > > N/A if there is reading error. > > > > I'm sending simple patch which just report NaN value when > > kernel report error. And also sensord will continue to > > reading next hwmon interfaces. > > > > Without this patch I see only these lines logged: > > > > sensord: Chip: radeon-pci-0100 > > sensord: Adapter: PCI adapter > > sensord: Error getting sensor data: radeon/#0: Can't read > > sensord: sensor read error (-1) > > > > And all other sensors are ignored and nothing more is > > logged. > > > > After my patch all hwmon sensors data are logged: > > > > sensord: Chip: radeon-pci-0100 > > sensord: Adapter: PCI adapter > > sensord: temp1: nan C > > sensord: Chip: i8k-virtual-0 > > sensord: Adapter: Virtual device > > sensord: Left Fan: 2685 RPM > > sensord: CPU: 49.0 C > > sensord: temp2: 55.0 C > > sensord: temp3: 39.0 C > > sensord: Chip: coretemp-isa-0000 > > sensord: Adapter: ISA adapter > > sensord: Physical id 0: 50.0 C > > sensord: Core 0: 47.0 C > > sensord: Core 1: 50.0 C > > BUMP! > > Can somebody review my patch? PING. -- Pali Rohár pali.rohar@xxxxxxxxx
Attachment:
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors