Hi Jaromír, Le samedi 05 mai 2012 à 19:40 +0200, Jaromír Červenka a écrit :

Hello,

today, one of my servers frozen out of nowhere. After hard reset, I found this in /var/log/messages

http://paste.opensuse.org/83089371

Configuration: openSUSE 12.1 Linux onca 3.1.9-1.4-default #1 SMP Fri Jan 27 08:55:10 UTC 2012 (efb5ff4) x86_64 x86_64 x86_64 GNU/Linux

Could it be a bug of kernel module or HW error?

That would be a hardware error. What would be interesting to know is what happened right before the SMBus transaction timeout. No hint in the logs? I can help you more if you provide additional information on your system. In particular : * Does the w83795 driver return sane values usually? * Does your board support IPMI, and do you have any ipmi driver loaded? * Is there a BMC connected to the machine? One frequent issue with the w83795 driver issue is when the monitoring device is accessed both by the w83795 driver and by IPMI (either from the host or from a BMC). As there is no synchronisation between them, it's only a matter of time before a collision happens and everybody gets confused. Another possibility is that you have another chip on the SMBus which went wrong and this blocked the SMBus completely, preventing other chips on the bus from being accessed. It may be possible to improve the i2c-i801 driver to recover from such situations. The problem is that it's not easy to differentiate between the bus being genuinely busy because another master is driving a transfer, and the bus looking busy because a slave is wrongly holding the lines low. Also, if a slave really went mad, there is nothing i2c-i801 can do, it's up to the slave's driver to reset the chip if it has a way to do so (but the hardware may be such that there is no such way other than a reboot.) -- Jean Delvare Suse L3 -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org