From owner-freebsd-hackers Tue Mar 21 22:52:02 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id WAA24606 for hackers-outgoing; Tue, 21 Mar 1995 22:52:02 -0800 Received: from vmbb.cts.com (vmbb.cts.com [192.188.72.18]) by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id WAA24600 for ; Tue, 21 Mar 1995 22:52:02 -0800 Received: from io.cts.com by vmbb.cts.com with smtp (Smail3.1.28.1 #9) id m0rrKGn-0000N4C; Tue, 21 Mar 95 22:51 PST Received: (from mdavis@localhost) by io.cts.com (8.6.11/8.6.9) id WAA01570 for hackers@freebsd.org; Tue, 21 Mar 1995 22:51:59 -0800 From: Morgan Davis Message-Id: <199503220651.WAA01570@io.cts.com> Subject: sio overruns To: hackers@FreeBSD.org Date: Tue, 21 Mar 1995 22:51:59 -0800 (PST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 662 Sender: hackers-owner@FreeBSD.org Precedence: bulk FYI: I've got a device connected to a serial port (a BBS) that requires DCD up all the time. However, every now and then a caller connects to the port via cu and drops carrier. The cu processes don't go away, and when the BBS sends data to the port, it triggers sio overrun errors. These overruns cause FreeBSD 2.1d0 to freak out (a technical term that means that pulling a ps will lock up the machine). I can access /proc, but a ps will hang that process. I can get logged in elsewhere, but it's tough to find PIDs to kill the hung sessions. All you can really do is down the machine. I didn't have this problem until I hooked up a BBS to the serial port.