On Thu, Nov 27, 2008 at 3:41 PM, Marco Chiappero <marco at absence.it> wrote: >> Hum, anything in dmesg? >> Anyway I'm attacching the new version (plus patch for ACPI - diffed >> against git-current, but it's easily adapted to older kernels). [...] > I compiled and loaded the new atk driver and compiled again lm-sensors with > the old patch (that was laying around in my /usr/src) and this time I > managed to see the atk driver readings from the "sensors" output (don't know > why it didn't work the first time, sorry). However I noticed that the code > you posted recently it's a little bit different, isn't? Should I try that > one too? The patch for lm-sensors? I've just rebased to SVN-current. > About the atk, it seems to work fine, but I haven't got the time to verify > the presence of strange issues, regarding for example suspend and/or > hibernate, stuff like that (does it matter?). The driver is stateless, so it shouldn't matter (testing won't harm though :P) > The only thing to say is that in the new code now shows all the readings in > the /sys/class/hwmon/hwmonX/ dir instead of the > /sys/class/hwmon/hwmonX/device/ I used to see with many other sensors chips. Both are supported, I'm not sure which one is "better" - I used the class directory since the ATK device also controls stuff other than hwmon. > I know that no fan control is possible with this driver, however I was > asking to myself whether it is possible to select on the fly one of the > three "silent, normal, turbo" profiles available in the BIOS settings. Even > if it is not a fine fan tuning, such regulation should be enough for me. It's doable using the Q-FAN interface. You use the interface calling GITM/SITM with class 0x4 (see the documentation embedded in my driver). There are at least 3 different types of control in this class, you'll have to reverse engineer the exact calling interface (start with the getters). > One, last thing: I use to collect and plot system data with collectd and > rrdtool, but collectd seems to ignore the atk driver. I know, I should read > the code, however any idea for addressing this issue? I suppose that collectd is using the unpatched libsensor which ignores the ATK device (it's attached to the unknown ACPI bus). Luca
