Re: [LinuxBIOS] [PATCH] adm1027.c fan control

[Ward Vandewege]

On Fri, May 11, 2007 at 02:25:08PM -0700, yhlu wrote:
> how about other MB not connect all of FAN etc.?
> or they will need other threshold for PWM?

OK, fair enough. Perhaps it is better to keep this kind of thing mainboard
specific.

I'll work on moving the code to mainboard.c file.

Thanks,
Ward.

[LinuxBIOS] [PATCH] adm1027.c fan control

[Stefan Reinauer]

* Ward Vandewege  [070511 23:35]:
> On Fri, May 11, 2007 at 02:25:08PM -0700, yhlu wrote:
> > how about other MB not connect all of FAN etc.?
> > or they will need other threshold for PWM?
>
> OK, fair enough. Perhaps it is better to keep this kind of thing mainboard
> specific.
>
> I'll work on moving the code to mainboard.c file.

Does it make sense to have an API letting the mainboard specific code
interface the component? Maybe even setting the FANs and threshold in
the config.lb file?

[LinuxBIOS] [PATCH] adm1027.c fan control

[yhlu]

On 5/12/07, Stefan Reinauer  wrote:
> Ward Vandewege  [070511 23:35]:
>
> Does it make sense to have an API letting the mainboard specific code
> interface the component? Maybe even setting the FANs and threshold in
> the config.lb file?

that will be good. to make the i2c device take register take param...

YH

[LinuxBIOS] [PATCH] adm1027.c fan control

[Ward Vandewege]

On Sat, May 12, 2007 at 02:33:55PM +0200, Stefan Reinauer wrote:
> Ward Vandewege  [070511 23:35]:
> > On Fri, May 11, 2007 at 02:25:08PM -0700, yhlu wrote:
> > > how about other MB not connect all of FAN etc.?
> > > or they will need other threshold for PWM?
> >
> > OK, fair enough. Perhaps it is better to keep this kind of thing mainboard
> > specific.
> >
> > I'll work on moving the code to mainboard.c file.
>
> Does it make sense to have an API letting the mainboard specific code
> interface the component? Maybe even setting the FANs and threshold in
> the config.lb file?

I think that could be very helpful. What kind of syntax do you have in mind?

It could potentially get a bit complicated; the Tyan s2881 for instance has 2
different HWM chips, each driving a number of fans and hooked up to different
temperature sensors.

Thanks,
Ward.

[LinuxBIOS] [PATCH] adm1027.c fan control

[Marc Jones]

Ward Vandewege wrote:
> On Sat, May 12, 2007 at 02:33:55PM +0200, Stefan Reinauer wrote:
> > Ward Vandewege  [070511 23:35]:
> > > On Fri, May 11, 2007 at 02:25:08PM -0700, yhlu wrote:
> > > > how about other MB not connect all of FAN etc.?
> > > > or they will need other threshold for PWM?
> > >
> > > OK, fair enough. Perhaps it is better to keep this kind of thing mainboard
> > > specific.
> > >
> > > I'll work on moving the code to mainboard.c file.
> >
> > Does it make sense to have an API letting the mainboard specific code
> > interface the component? Maybe even setting the FANs and threshold in
> > the config.lb file?
>
> I think that could be very helpful. What kind of syntax do you have in mind?
>
> It could potentially get a bit complicated; the Tyan s2881 for instance has 2
> different HWM chips, each driving a number of fans and hooked up to different
> temperature sensors.
>
> Thanks,
> Ward.

I agree with Ward. Config.lb is complicated enough. I would rather see a 
  user setup function that gets called from mainboard.c with the 
appropriate thresholds.
Marc

[LinuxBIOS] [PATCH] adm1027.c fan control

[yhlu]

On 5/14/07, Marc Jones  wrote:
> Ward Vandewege wrote:
> > > > I'll work on moving the code to mainboard.c file.
> > >
> > > Does it make sense to have an API letting the mainboard specific code
> > > interface the component? Maybe even setting the FANs and threshold in
> > > the config.lb file?
> >
> > I think that could be very helpful. What kind of syntax do you have in mind?
> >
> > It could potentially get a bit complicated; the Tyan s2881 for instance has 2
> > different HWM chips, each driving a number of fans and hooked up to different
> > temperature sensors.
> >
> > Thanks,
> > Ward.
>
> I agree with Ward. Config.lb is complicated enough. I would rather see a
>   user setup function that gets called from mainboard.c with the
> appropriate thresholds.

OK, ward could work on put that on mainboard.c at first. and late
someone could optimize it ..

YH

[LinuxBIOS] [PATCH] adm1027.c fan control

[ron minnich]

I recomment we save such good ideas for V3. I am going to start on my
first real V3 mainboard this week -- an LX board -- and we can start
getting real-world experience now.

thanks

ron

[LinuxBIOS] [PATCH] revised fan control for tyan s2881

[Ward Vandewege]

Hi all,

Please find the revised fan control patch for the Tyan s2881 attached.
Everything is now in the tyan/s2881/mainboard.c file.

It works, but the implementation is a bit silly with the dummy PNP device I
had to create. I'd greatly appreciate comments to make this code more
elegant.

The mainboard.c file did not have a copyright header; I've added it. Yinghai,
I've assumed you were the author of the original 5-line file, and put AMD as
a copyright holder. I've also assumed you wrote it in 2005 since I know it's
been around at least since then. Is that correct?

Suggestions are of course welcome.

Thanks,
Ward.

[LinuxBIOS] [PATCH] revised fan control for tyan s2881

[yhlu]

I made some change, please check it

Acked_by: Yinghai Lu

YH

On 5/16/07, Ward Vandewege  wrote:
> Hi all,
>
> Please find the revised fan control patch for the Tyan s2881 attached.
> Everything is now in the tyan/s2881/mainboard.c file.
>
> It works, but the implementation is a bit silly with the dummy PNP device I
> had to create. I'd greatly appreciate comments to make this code more
> elegant.
>
> The mainboard.c file did not have a copyright header; I've added it. Yinghai,
> I've assumed you were the author of the original 5-line file, and put AMD as
> a copyright holder. I've also assumed you wrote it in 2005 since I know it's
> been around at least since then. Is that correct?
>
> Suggestions are of course welcome.
>
> Thanks,
> Ward.

[LinuxBIOS] [PATCH] revised fan control for tyan s2881

[Ward Vandewege]

On Wed, May 16, 2007 at 04:21:11PM -0700, yhlu wrote:
> I made some change, please check it

Great. Your changes make sense. I'll test them on real hardware Friday, but
this looks good to me. Thanks for clarifying the licensing situation, too.
> Acked-by: Yinghai Lu

Acked-by: Ward Vandewege

Thanks,
Ward.

[LinuxBIOS] [PATCH] revised fan control for tyan s2881

[Ward Vandewege]

On Wed, May 16, 2007 at 04:21:11PM -0700, yhlu wrote:
> I made some change, please check it
>
> Acked-by: Yinghai Lu

Acked-by: Ward Vandewege

I've fixed a small issue with your changes (struct bus children vs. child)
and tested the revised patch on real hardware. It works fine. Could someone
apply the attached patch to the tree?

Thanks,
Ward.

[LinuxBIOS] [PATCH] revised fan control for tyan s2881

[Uwe Hermann]

On Fri, May 18, 2007 at 11:59:27AM -0400, Ward Vandewege wrote:
> Acked-by: Ward Vandewege

For the record: committed in r2674.

Uwe.