Date: Sat, 22 Jul 95 23:30:05 MDT From: terry@cs.weber.edu (Terry Lambert) To: ache@astral.msk.su (=?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?= aka) Cc: hackers@freebsd.org, jkh@violet.berkeley.edu, bde@zeta.org.au, harry@hgac.com Subject: Re: dial up at > 9600 baud Message-ID: <9507230530.AA24940@cs.weber.edu> In-Reply-To: <pWyhR4m4e2@astral.msk.su> from "=?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?= aka" at Jul 23, 95 06:59:08 am
next in thread | previous in thread | raw e-mail | index | archive | help
> >PPP always connected at 9600 baud. Found that in getty.c the open() function > >blocks until carrier detected. This happens before the port is set up using > >the getty args. Changed getty to use non-blocking open(), then fall through > >to set up the port with getty args, and then wait for carrier using ioctl() > >similar to slattach. See attached diff file for getty.c in FreeBSD 2.0.5. > > I don't think, that waiting for carrier is right way. > Better way will be to open device as non-blocked, > then setup all tty flags (incl. speed), then reopen device > in blocked mode just before reading anything. > > BTW, Bruce, is it possible now to switch to blocked > mode without reopening? No, it's not possible, since we are talking about whether the open flag is set or not for subsequent opens. The reason the ioctl() loses is taht it will prevent bidirection use of the port, with the getty open always having succeeded. I don't think I quite understand why the previous behaviour was a problem in the first place -- perhas the original poster could explain the rationale for his changes a bit better? Terry Lambert terry@cs.weber.edu --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9507230530.AA24940>