- From: Frank Mori Hess <fmhess_at_users.sourceforge.net>
- Date: Tue, 11 Mar 2003 13:59:59 -0600
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Tuesday 11 March 2003 03:50 am, Örnólfur E. Rögnvaldsson wrote: > Hi! > > I have some code that uses comedi to do continuous sampling from an NI > 6023E board. It has worked fine so far but now I've experienced an "ai > error a_status=8839" condition on a single occasion (after ~3hrs on one > system, another identical system runs w/o problems for 3 days). The problem > is not so much that the error occurs, if only I knew how to recover from it > (even if that simply means restarting the program:-) but apparently the > code hangs after the ai error. From looking at the ni_mio_common code I > guess the error is triggered during interrupt processing? Ideally I would > like to be notified of the error by my > ret=read(comedi_fileno(dev),buf,bufsz) call returning -1, but that seems > not to happen here? How do I get notified of the error? Hmmm, maybe my code > hangs due to some of my own bugs, but it's a bit suspicious that this > coincides with the ai error? > > Versions used: > comedi-0.7.65 > comedilib-0.7.19 > Catching the error should be caught in cvs now. If you want to fix 0.7.65, just add a call to comedi_event() like comedi_event(dev,s,s->async->events); right before the interrupt handler returns after detecting the error. - -- Frank -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE+bkA/5vihyNWuA4URAldTAJsGD5xpCX2aQe+QzApCNIioQVv0UgCfQavr YD6gmqnjQPhFcrYuaWwMJs4= =i8EG -----END PGP SIGNATURE-----
Received on 2003-03-11Z19:59:59