On Wednesday 14 January 2004 05:32, Mark Studebaker wrote:
> The names of clients and the addresses of their subclients are "well known"
> so in prog/sensors/main.c we just look for them (see that file for more
> info). Your proposal to look for the string "subclient" will work equally
> well. Feel free to submit a sensord patch to us using either technique.
If I understood you correctly (subclient does not provide this information)
here is patch. It simply ignores any chip that has a space in name. To my
best knowledge every chip has single name without spaces except for special
case of subclients (at least that what I have done in my kernel patch). If
this is not true any more - tell so and I change it to look for "subclient"
substring.
regards
-andrey
> mds
>
> Greg KH wrote:
> > On Sun, Jan 11, 2004 at 01:19:18PM +0300, Andrey Borzenkov wrote:
> >>I do not know if it is a sensord (should not attempt to read subclient)
> >> or kernel (does not provide sufficient information) problem. Subclient
> >> can be determined by its name:
> >>
> >>{pts/1}% for i in /sys/bus/i2c/devices/*
> >>do
> >>echo ${i:t}:$(<$i/name)
> >>done
> >>0-002d:as99127f
> >>0-0048:as99127f subclient
> >>0-0049:as99127f subclient
> >>
> >>Greg, this is supposed to be established convention now?
> >
> > I think so, yes. I'll let the userspace people answer your other
> > questions.
> >
> > thanks,
> >
> > greg k-h
-------------- next part --------------
A non-text attachment was scrubbed...
Name: lm_sensors-2.8.2-subclient.patch
Type: text/x-diff
Size: 570 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20040118/67b1eafe/attachment.bin
