On Sun, August 2, 2020 8:54 PM, Pavel Machek wrote: > Hi! > > > > +static DEVICE_ATTR_RW(ps_seq_id); > > > +static DEVICE_ATTR_RW(ps_seq_loop); static > > > +DEVICE_ATTR_RW(gpi_seq_id0); static DEVICE_ATTR_RW(gpi_seq_id1); > > > +static DEVICE_ATTR_RW(gpi_seq_id2); static > > > +DEVICE_ATTR_WO(patterns); > > > > Should this be a binary attribute instead of having string parsing in > > the kernel? > > Plus... is this suitable userland interface? Should it be documetned? It could be used by sysfs interface and string parsing, and I added some examples ahead of id and loop attributes, but for the patterns yes, it should be. As I have already added some comment before for the first comment from Dmitry, now the upload effect covers this attributes, so I would like to remove them and I have added some code a little bit more in the upload effect function in the next patch. Thanks for your comments. > > -- > (english) http://www.livejournal.com/~pavelmachek > (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html Kind regards Roy
