Hi Andrew, On Jan 30, 2008 5:18 AM, Andrew Voznytsa <andrew.voznytsa at gmail.com> wrote: > Hi Juerg, > > Of course - works perfectly. It said > > [..] > Trying family `SMSC'... Yes > Found `SMSC SCH5027D-NW Super IO' > (hardware monitoring capabilities accessible via SMBus only) > [..] There should be another instance where the SMSC is detected as an I2C device. Can you post the complete log? > If I can do anything else - just let me know. > > Should I try some lm-sensors's svn branch/revision to test if all fans are > monitored now? If there is such - could you send me URL? That won't help. I need to update the dme1737 driver first which is part of the kernel, not lm-sensors. ...juerg > Many thanks, > Andrew Voznytsa > > > > > > -----Original Message----- > > From: Juerg Haefliger [mailto:juergh at gmail.com] > > Sent: Wednesday, January 30, 2008 7:22 AM > > To: Andrew Voznytsa > > Cc: lm-sensors at lm-sensors.org > > Subject: Re: SMSC Super I/O: unknown chip with ID 0x8902 > > (was: lm85 in Intel S3000AH case) > > > > Hi Andrew, > > > > Could you run the attached script and post the output? It should > > correctly identify the SCH5027. > > > > ...juerg > > > > > > On Jan 18, 2008 7:25 AM, 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. > > > > > > > > > > Just let me know if I can help with testing - I can't give access to my > > > S3000AH-based system but definitely can run some tests, etc. > > > > > > All the best, > > > Andrew > > > > > > > > > > >
