- From: frank mori hess <fmhess_at_students.uiuc.edu>
- Date: Mon, 8 Jul 2002 19:43:39 -0500 (CDT)
On Mon, 8 Jul 2002, David Schleef wrote: > On Mon, Jul 08, 2002 at 07:29:38PM -0500, frank mori hess wrote: > > > > Dave, > > > > It seems to me that comedi_lock() in kcomedilib is currently broken, due > > to the fact that the comedi_t values are not unique. Two processes that > > open the same device will receive as their comedi_t pointer the same > > address to the same comedi_device. > > This is the expected behavior. > I don't understand. Isn't comedi_lock() useless then? The if(s->lock && s->lock!=d) will always be false. Frank
Received on 2002-07-08Z23:43:39