Messages in this thread | | | Subject | Re: [PATCH 2/2] enhance /sys/class/rfkill/<rfkill>/state interface | From | Marcel Holtmann <> | Date | Sun, 21 Feb 2010 11:07:35 +0100 |
| |
Hi Florian,
> > > Introduce a new state-value RFKILL_STATE_SOFT_AND_HARD_BLOCKED > > > which is returned only through the sysfs state file. > > > The other interfaces are designed so that they don't need this extra > > > state. > > > > > > This allows the sysfs to represent all possible states an rfkill > > > driver can > > > have. > > > > > > Signed-off-by: Florian Mickler <florian@mickler.org> > > > --- > > > > > > After stumbling over this arbitrary limitation of > > > sys/class/rfkill/*/state I > > > wondered what would hinder this patch? > > > > This is not backward compatible, so can't be done. > > > > johannes > > hmm... ah, i see... if driver is in hard'n'soft-block state > an userspace program would expect to read hardblock instead of the > new hard'n'softblock-state... > now that i think of it, it even becomes obvious :)
a userspace program would be expected to use /dev/rfkill and do this properly. Don't bother with sysfs at all.
Regards
Marcel
| |