Hi Andrew, On Jan 18, 2008 3:23 PM, Andrew Voznytsa <andrew.voznytsa at gmail.com> wrote: > > Hi Juerg, > > > > > > > > Trying family `SMSC'... Yes > > > > > > > > > > > > > > Found unknown chip with ID 0x8902 > > > > > > > > > > > > This is indeed unknown. Can you check the markings on the chip so > > that > > > > > > we know what we're dealing with? From the product picture my guess > > is > > > > > > it's the chip in the QFP package close to the IDE connector. It > > should > > > > > > read something like SMSC xyz. > > > > > > > > > > There is chip marked SMSC SCH-5027D(not sure that D, might be 0)-NW > > > > > > > > > > According to spec which comes with MB it should be SMSC SC-5027D > > > > > > > > Can you run > > > > i2cdetect -y 0 > > > > > > > > > > > > > Of course - here is output: > > > > > > 0 1 2 3 4 5 6 7 8 9 a b c d e f > > > 00: -- -- -- -- -- 08 -- -- -- -- -- -- -- > > > 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- > > > 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- UU -- > > > 30: 30 31 32 33 -- -- -- -- -- -- -- -- -- -- -- -- > > > 40: -- -- -- -- 44 -- -- -- -- -- -- -- -- -- -- -- > > > 50: 50 51 52 53 -- -- -- -- -- -- -- -- -- -- -- -- > > > 60: -- 61 -- 63 64 -- -- -- -- 69 -- -- -- -- -- -- > > > 70: -- -- -- -- -- -- -- -- > > > > OK, I figured it out, I believe :-) I received the datasheet for the > > SCH5027 from SMSC and after a quick glance at it, it looks like the > > hardware monitoring features of the DME1737, SCH5027 and EMC6D10x are > > identical but they advertise themselves differently. In your case, the > > chip is detected as a EMC6D10x and sensors-detect suggests the lm85 > > driver. However, I believe the dme1737 driver is more appropriate. I > > will take a closer look at the various datasheets over the next couple > > of days to confirm my theory. In the meantime, you can try the > > following: > > 1) remove the lm85 driver (rmmod lm85) > > 2) force the dme1737 driver (modprobe dme1737 force=0,0x2e) > > 3) post the dme1737-related messages in /var/log/messages to the mailing > > list > > Here is output: > > Jan 18 16:09:12 farm dme1737 0-002e: Failed to query Super-IO for optional > features. Ah yes, the dme1737 driver doesn't know anything about the SCH5027 Super-IO so it can't determine if fans 5&6 and PWMs 5&6 are configured. > Jan 18 16:09:12 farm dme1737 0-002e: Optional features: pwm3=yes, pwm5=no, > pwm6=no, fan3=yes, fan4=yes, fan5=no, fan6=no. If the BIOS shows 6 fans, fans 5&6 should become available once the driver is fixed to recognize the SCH5027. > Jan 18 16:09:12 farm i2c-adapter i2c-0: Found a DME1737 chip at 0x2e (rev > 0x00) > > > > > Note that the SCH5027 only supports 4 fans, so you won't see fan5&6. > > How many fans do you have in the machine? Can you see RPMs for all of > > them in the BIOS? Correction: This is not true. The SCH5027 supports 6 fans, just like the dme1737. > I've 2 fans - CPU (fan1) and chassis (fan5). They are connected as Intel's > manual says. > > BIOS shows 6 fans - it shows some (reasonable) RPMs for fans1, 5 and zero > for rest (nothing attached there). BIOS is able to control RPMs itself, > depending on CPU, chassis and DIMM temperature sensors. Again, I'll take a look at the datasheets and if the chips are indeed compatible, I'll fix the dme1737 driver to support the sch5027. I don't have access to a test machine until next week, so it has to wait until then. ...juerg > Many thanks for looking on it, > Andrew > > >