Date: Tue, 14 Apr 1998 15:50:50 -0700 From: Mike Smith <mike@smith.net.au> To: David McNett <nugget@slacker.com> Cc: hackers@FreeBSD.ORG, mike@smith.net.au Subject: Re: Indus Drivers Message-ID: <199804142250.PAA01253@dingo.cdrom.com> In-Reply-To: Your message of "Tue, 14 Apr 1998 14:18:12 CDT." <19980414141812.09510@slacker.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> On 14-Apr-1998, Mike Smith wrote: > > Try the USR Courier I-modem. > > I have two of these, and have my hands near another one. In all cases, > I have had horrific luck getting FreeBSD to play well with the courier. > This is using byterunner serial ports at 230.4kbps. Have you tried the internal units? I have one here but it's the International model, and incapable of operating in the USA. > Under periods of > heavy inbound traffic, the courier will simply cease processing inbound > traffic altogether. No traffic, no blinkenlights, nothing. The only > solution I've found when this happens is to powercycle the courier. > (This on even the most recent courier i-modem firmware from last week) Bleagh. > Not that I hold FreeBSD at fault for this, my gut reaction is that it > is a problem with the courier. I can reproduce the exact same behavior > running the same hardware on a Linux box with non-16650 aware > setserial. Linux (RH 4.2 to be exact). However, with a more recent > setserial (16650 aware) produces a very solid link that does not > exhibit this behavior. This makes it sound like it's a flow-control related issue. Have you tried chaging the receive FIFO settings under FreeBSD? Have you put a serial tester on the line and looked at the status of the flow control signals when the USR unit locks up? Have you pestered USR about this? > I would love nothing more than to ditch my linux boxes in favour of > FreeBSD routers, but do not care to spend any money on additional > hardware to accomplish this. > > I bought a gtek 16554 board to experiment with replacing the byterunner > board, but have since realized that there are no FreeBSD drivers for > it. The only 16554 I know of is a 4-port 16550 clone made by Startech. It looks just like 4 separate 16550's. > Serial i/o is a black art at this level, and I'm well beyond my > skillset. Anyone have any recommendations? What is in sio.c that > could be triggering this behavior, and what's magic about the newer > linux setserial code that would somehow rectify the problem? Differet flowcontrol behaviour is my guess. If you talk to the USR weenies and they complain about not having a test system/not being able to reproduce the problem, it might be possible to send them a suitably configured system to play with. -- \\ Sometimes you're ahead, \\ Mike Smith \\ sometimes you're behind. \\ mike@smith.net.au \\ The race is long, and in the \\ msmith@freebsd.org \\ end it's only with yourself. \\ msmith@cdrom.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199804142250.PAA01253>