John Ferlito wrote: > Also can the user cat /dev/ttyS0 ?
Hmm, yes it can. Now's when I probably need to provide some more info. The process in question is a CUPS [0] backend driver that talks to a serial printer. The process is started by CUPS with a whole bunch of special enviroment variables and command line parameters which makes it a little difficult to debug. By writing to a log file I know that when the process is started by CUPS it has a user and group id of "lp" and that "lp" is a member of the "dialout" groups. Su-ing to the "lp" user account and catting to the serial port works as does fiddling the right command line and environment to run the backend directly from the "lp" account. That means that CUPS may be doing something to the process before it execs the backend driver, possibly using capabilities [1]. More debugging needed. Cheers, Erik [0] http://www.cups.org/ (I know is about the best common printing system available for Linux but damn its bloody awful!) [1] http://www.securityfocus.com/infocus/1400 -- ----------------------------------------------------------------- Erik de Castro Lopo ----------------------------------------------------------------- "... so the notion that it is meaningful to pass pointers to memory objects into which any random function may write random values without having a clue where they point, has _not_ been debunked as the sheer idiocy it really is." -- Erik Naggum, comp.lang.lisp _______________________________________________ coders mailing list coders@slug.org.au http://lists.slug.org.au/listinfo/coders
