[lm-sensors] SMSC Super I/O: unknown chip with ID 0x8902 (was: lm85 in Intel S3000AH case)

Juerg Haefliger juergh at gmail.com
Wed Jan 30 17:08:17 CET 2008


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: [lm-sensors] 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
> > >
> > >
> > >
>
>




More information about the lm-sensors mailing list