[superiotool] patch for fintek f71889fg

[David Hendricks]

The patch (attached) was tested by a user on IRC who had the F71889FG. I
wrote it using documentation from Fintek's website available here:
http://www.fintek.com.tw/files/productfiles/F71889_V0.28P.pdf

This patch also seems to work for the F71889ED, which uses 0x09 and 0x09 for
chip ID bytes 1 & 2. However, I have not been able to find documentation to
verify that the two chips are identical from superiotool's perspective.

The F71889 seems popular on current generation platforms with AMD chipsets,
in case there are folks looking to try it on a presently unsupported
board...

Signed-off-by: David Hendricks

[Uwe Hermann]

Hi,

On Mon, Nov 01, 2010 at 01:02:53PM -0700, David Hendricks wrote:
> The patch (attached) was tested by a user on IRC who had the F71889FG. I
> wrote it using documentation from Fintek's website available here:
> http://www.fintek.com.tw/files/productfiles/F71889_V0.28P.pdf
>
> This patch also seems to work for the F71889ED, which uses 0x09 and 0x09 for

Both times 0x09? Or is this a typo?
> chip ID bytes 1 & 2. However, I have not been able to find documentation to
> verify that the two chips are identical from superiotool's perspective.
>
> The F71889 seems popular on current generation platforms with AMD chipsets,
> in case there are folks looking to try it on a presently unsupported
> board...
>
> Signed-off-by: David Hendricks

Are you sure this is the correct patch? It doesn't seem to match the
datasheet in a number of places, e.g. 0x20 and 0x21 (IDs) are incorrect
in NOLDN (as well as most other values in NOLDN), some registers are missing
completely, some LDNs are missing completely etc.

Is this for another Super I/O, or remainders of copying another table?

Uwe.

[David Hendricks]

On Tue, Nov 2, 2010 at 12:45 PM, Uwe Hermann  wrote:
> Hi,
>
> On Mon, Nov 01, 2010 at 01:02:53PM -0700, David Hendricks wrote:
> > The patch (attached) was tested by a user on IRC who had the F71889FG. I
> > wrote it using documentation from Fintek's website available here:
> > http://www.fintek.com.tw/files/productfiles/F71889_V0.28P.pdf
> >
> > This patch also seems to work for the F71889ED, which uses 0x09 and 0x09
>
> for
>
> Both times 0x09? Or is this a typo?

FWIW, lm-sensors has these two entries:
}, {
 name => "Fintek F71889FG Super IO Sensors",
 driver => "f71882fg",
 devid => 0x0723,
 logdev => 0x04,
 features => FEAT_IN | FEAT_FAN | FEAT_TEMP,
 }, {
 name => "Fintek F71889E Super IO Sensors",
 driver => "to-be-written",
 devid => 0x0909,
 logdev => 0x04,
 features => FEAT_IN | FEAT_FAN | FEAT_TEMP,
 }, {

I am not certain what the difference is between the two chips, if any
are discernible from superiotool's perspective. The F71889FG datasheet seems
to be the public one -- the 0x23 and 0x07 chip IDs match the documented
values. However, the chip I tested with has F71889ED printed on it and has
0x09 for the two chip ID bytes.
> > chip ID bytes 1 & 2. However, I have not been able to find documentation
>
> to
> > verify that the two chips are identical from superiotool's perspective.
> >
> > The F71889 seems popular on current generation platforms with AMD
>
> chipsets,
> > in case there are folks looking to try it on a presently unsupported
> > board...
> >
> > Signed-off-by: David Hendricks
>
> Are you sure this is the correct patch? It doesn't seem to match the
> datasheet in a number of places, e.g. 0x20 and 0x21 (IDs) are incorrect
> in NOLDN (as well as most other values in NOLDN), some registers are
> missing
> completely, some LDNs are missing completely etc.
>
> Is this for another Super I/O, or remainders of copying another table?

I took a glance at the patch again and agree that there are several places
where the code does not match the doc. Probably sloppy copy + paste on my
part. Let's hold off on this patch until NTU, myself, or someone has time to
check the accuracy.

If someone happens to have a F71889*ED* datasheet with the chip ID bytes
both listed as 0x09, that would help.

[Uwe Hermann]

On Thu, Nov 04, 2010 at 12:46:16PM -0700, David Hendricks wrote:
> I am not certain what the difference is between the two chips, if any
> are discernible from superiotool's perspective. The F71889FG datasheet seems
> to be the public one -- the 0x23 and 0x07 chip IDs match the documented
> values. However, the chip I tested with has F71889ED printed on it and has
> 0x09 for the two chip ID bytes.
> I took a glance at the patch again and agree that there are several places
> where the code does not match the doc. Probably sloppy copy + paste on my
> part. Let's hold off on this patch until NTU, myself, or someone has time to
> check the accuracy.

I fixed up all issues I could see and committed a merged version from
your patch and the one from NTU (kept both Signed-off-by's) in r6131.

If you have a board with the 0909 ID you can test on, feel free to post
a patch which adds at least detection support for that chip. Until we
have a datasheet to check the registers I'd rather not add any guessed
register values to superiotool, though.

NTU, can you please post a "superiotool -d" output (use the latest
superiotool from trunk) from your board?

Thanks, Uwe.