Hello Luca, by googling I found this discussion on the sensors-mailinglist (so maybe some CC: are lost), and I observe the same problem (comments below) >On Fri, Dec 11, 2009 at 7:03 PM, <comsublant <at> gmx.net> wrote: >> On Fri, Dec 11, 2009 at 02:04:52PM +0100, Luca Tettamanti wrote: >>> It's very strange since I had a tester that helped my with the very same >>> board. >> >> This might be mistakable: this is a pure p7p55d, not -premium or >> something. > > I have a report of a working "pro" board, shouldn't change much, at > least as far as the ACPI interface is concerned. > >>> What kind of keyboard/mouse are you using? PS2 or USB? >> >> PS2 >> >>> Can you access the machine remotely (ssh)? If so, please send me a >>> dump of /proc/interrupts _after_ the keyboard becomes unresponsive. >> >> no remote access at the moment :( > > Ok, does it respond to the power button? A clean shutdown may help to > preserve something interesting in the logs. > >>> Ok, the only difference between working and non-working is the >>> atk0110 driver? >> >> yes >> >>> Locate atk_ec_present function in asus_atk0110.c; put a >>> >>> return 0; >>> as first instruction; recompile and see if it still locks up the keyboard. >>> Also enable HWMON_DEBUG_CHIP. >> >> done. >> its not locking up anymore, > [...] >> its not changing though. > > This is expected, since the EC is responsible of updating the values. > The behaviour that you describe seems to indicate a runaway interrupt > (possibly from the EC). > Please send me a copy of the DSDT (/sys/firmware/acpi/tables/DSDT) and > the output of lspci -vn > > thanks, > Luca in my situation: just a few minutes into a KDE session (doing nothing extraordinary) first the mouse becomes unresponsive, and about 15 seconds later also the keyboard (toggling num-lock does nothing). Both are connected via PS. This has happened with 2.6.33-rc5 and rc6 as taken from the opensuse repository. I was thinking that maybe the X-server is doing something wrong, and for both kernels, I had renamed the module "nouveau.ko" to something else (and also the libdrmnouveau files), and the error has not yet come back (but I need more testing). I don't know how this is related to the atk0110, but at least sensors-detect finds some i2-c devices on the nvidia card (and can not do much with them). My board is a P7P55D, and I also obtain these deep temperatures: [...] CPU Temperature: +32.5°C (high = +45.0°C, crit = +45.5°C) MB Temperature: +31.0°C (high = +45.0°C, crit = +46.0°C) [...] Is there a chance to change the "high" and "crit" values to something higher ? Because the chassis fan is extremely loud ... best regards, Michael _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
