Date: Sat, 3 Jul 1999 19:57:33 -0500 (CDT) From: Chris Dillon <cdillon@wolves.k12.mo.us> To: Marcel Moolenaar <marcel@scc.nl> Cc: stable@FreeBSD.ORG Subject: Re: Linux emulation Message-ID: <Pine.BSF.4.10.9907031934570.63435-100000@mail.wolves.k12.mo.us> In-Reply-To: <377DB035.7EE9AF75@scc.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 3 Jul 1999, Marcel Moolenaar wrote: > [snip] > > > > Andrij Korud wrote: > > > > "LINUX: 'ioctl' fd=0, typ=0x54(T), num=0x17 not implemented" > > [snip] > > > I wrote: > > > Can you apply the following patch and tell me if it solves your problem > > > without creating new problems? :-) > > [snip] > > Chris Dillon wrote: > > I just tried this patch on my server and it does the job, almost. > > There is no "unsupported ioctl" error message, but the daemon can't > > communicate with the UPS _unless_ I frob the port with tip(1) while > > the daemon is running. The exact moment that I tip the serial port, > > the daemon comes alive and tells me UPS communications were > > established and its happy. tip(1) isn't able to actually talk to the > > UPS, though it doesn't give me any errors of any kind, and I just ~. > > to exit out. This would be acceptable if I could find an automated > > way of frobbing the serial port so that the daemon works. :-) > > I figured something like that would do the trick. A ktrace showed that > read(2) on the com-port returns -1 with errno set to -11. Everything else > seemed to work. > > Can you tell us what kind of frobbing you did? Maybe there's a bug > somewhere in the TCGETS/TCSETS iotcls (these are used right after the > TIOCMBIC ioctl). In /etc/remote: smartups:dv=/dev/cuaa1:br#2400:pa=none: Then simply 'tip smartups' and immediately exit with ~. or kill it. > In the mean time I prepare the patch for being committed. > > As for automated frobbing :-) Maybe it helps to initialize the device at > boot-time. See /etc/rc.serial I'm not sure that /etc/rc.serial would help, but I'll play with it later and see. ... > I have my uname patch applied by default, which means that uname returns > "Linux" by default. When you run the pap_inst script in a Linux sh, it all > works out of the box. Changing the defaults and directories in the script > to match a FreeBSD environment is something which would be nice in case of > a real port though (hint :-) That's exactly what I did, and I will do a port as soon as I can get it working without needing to touch the serial port first. :-) In the event that messing with /etc/rc.serial does the trick, does anyone know of any ports that modify anything in /etc to use as an example? I'm not so sure that is an acceptable thing for a port to do, but if it is, it would make things a lot easier. -- Chris Dillon - cdillon@wolves.k12.mo.us - cdillon@inter-linc.net FreeBSD: The fastest and most stable server OS on the planet. For Intel x86 and Alpha architectures (SPARC under development). ( http://www.freebsd.org ) "One should admire Windows users. It takes a great deal of courage to trust Windows with your data." To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.9907031934570.63435-100000>