Date: Mon, 12 May 1997 11:40:17 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: luigi@labinfo.iet.unipi.it (Luigi Rizzo) Cc: terry@lambert.org, tom@sdf.com, brian@awfulhak.org, hackers@FreeBSD.ORG Subject: Re: sio flow control problems with 2.2? Message-ID: <199705121840.LAA07959@phaeton.artisoft.com> In-Reply-To: <199705120708.JAA21494@labinfo.iet.unipi.it> from "Luigi Rizzo" at May 12, 97 09:08:14 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > This is why there are stone tablets in a cave on Mt. Ararat which > > (loosely translated) say: > > > > "Thou shalt not set the port speed higher than the committed > > data rate for the communications device unless thine flow > > control is active and trustworthy". > > which was my question in the first place.. do I have working flow > control ? I just don't know how broken are my internal modems. In my > case, with a 28.8 modem, even stty speed 19200 causes the same failure. > That would suggest a problem at the receive side, except that I don't > know how constant is the modem speed over time... This is a very hard question to answer. Most modem hardware does not come in boxes with large red and white "Now With WORKING Flow Control!" stickers. Most modem manufacturers have screwed this up at one time or another, and they aren't about to admit to something which might force them to replace hardware across the board rather than replacing on a case-by-case (ie: cheaper) basis. When people started cloning Microcomm's MNP, they put much less RAM in their modems to save costs, and then *forced* the use of *in-band* flow control. And you think you have problems now... 8-(. What do you know about your modem? Does it have discrete UARTs, or is it doing an emulation? (It's probably an emulation). Which emulation chipset is it running, and what mask revision does it have? Once you are armed with the answers to these questions, it's possible to ask the manufacturer about data overrun. Are you running US Robotics on one end and Rockwell on the other? This combination is known to, for some older (than 6 months) US Robotics modems, especially the Fax modems, lock up the US Robotics chips. Why they are not compatible with Rockwell chipsets is unknown. As far as getting a failure with port speed 19.2 and modem speed 28.8... 28.8 is not a comitted data rate. They can not possibly guarantee that they will be aboe to send 2880 characters oer second -- it's a statistical value which assumes compressability of the transferred data. Any change you can got to an external modem? Barring that, any chance you can swap modems between the ends of your connection to see if the problem is bidirectional, or unidirectional? Regards, Terry Lambert terry@lambert.org --- 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?199705121840.LAA07959>