fscpos driver

Stefan Ott stefan at desire.ch
Wed Jan 19 03:15:34 CET 2005

On Wed, Jan 05, 2005 at 08:31:17PM +0100, Jean Delvare wrote:
> > > > It's odd that the min value would affect the speed, since it
> > > > supposedly is only a limit to the alarm condition, not a speed
> > > > target.
> > >
> > > According to the PDF you sent me, this really is meant to actually
> > > control the fan speed.
> > 
> > Wow. I'll take a look at it this evening. If this is true, then this
> > would correspond to either an analog output or PWM output, and
> > correspond to a file names pwm1, not fan1_min.
> I confirm this. The datasheet is rather clear about this, as you said -
> although it doesn't mention whether analog output amplitude modulation
> or PWM is used, nor does it states how temperature interacts with the
> final speed. The fact that temperature somehow matters may explain the
> strange behavior you noticed.
> Note that I suspect that fscher isn't correct on some points. It seems
> to create files named fanN_pwm (old standard) instead of pwmN (new
> standard), and also maps fanN_div to fan ripple registers, while these
> are two different things. I won't blame Reinhard Nissl too hard since I
> might be responsible for some of the errors, as I gave him possibly
> unenlightened advice while he did the work.

Okay. Meanwhile I changed the code so that the fan status can't be written into
anymore - it's automatically cleared when the speed raises above 0 after a
failure. However, I don't think I can do the same with the temperature and 
watchdog status as I can't really test them and I wouldn't want to clear a
temperature alarm without knowing that the chip takes care of setting it again
if required.

I also renamed the fanN_min to pwmN and still haven't figured out how they work,
that is, what the chip does with those values. Libsensors should probably know
about the pwm change and that the features 'rev' and 'control' are gone. Should
I create a patch or are you going to do that?

For me, the remaining issues are not really relevant, thus I'd say the driver
is ready now. What do you think?

I attached a patch against kernel 2.6.11-rc1, if you'd like to take a look at
the code.

Stefan Ott

"Give a man a fire and he's warm for a day, but set fire to him and he's
warm for the rest of his life."  -- Terry Pratchett, Jingo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: kernel-patch-fscpos-2.6.11-rc1-20050119.bz2
Type: application/octet-stream
Size: 5400 bytes
Desc: not available
URL: <http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20050119/e15f068b/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20050119/e15f068b/attachment.sig>

More information about the lm-sensors mailing list