From owner-freebsd-bugs Fri Mar 14 07:51:01 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA24493 for bugs-outgoing; Fri, 14 Mar 1997 07:51:01 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA24488 for ; Fri, 14 Mar 1997 07:50:58 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id CAA31096; Sat, 15 Mar 1997 02:50:29 +1100 Date: Sat, 15 Mar 1997 02:50:29 +1100 From: Bruce Evans Message-Id: <199703141550.CAA31096@godzilla.zeta.org.au> To: ahd@kew.com, bde@zeta.org.au, j@uriah.heep.sax.de Subject: Re: i386/2984: serial console speed goes to hell during boot Cc: freebsd-bugs@freefall.freebsd.org Sender: owner-bugs@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> > When booting the 2.2 GAMMA release using a serial console >> > the output (to an IBM PS/1 previously successfully used >> > for a serial console for 2.1.5-RELEASE for another 386), >> > output is normal 9600 bps output until about the time the >> > NPU is detected. Then, output slows to a crawl (~ 1 >> > character/second) until shutdown ... the last two lines >> > before reboot are printed at a normal speed (again), implying >> > the hardware is (still) fine. >> >> Check that nothing in /etc/rc.serial touches the console port. > >But rc.serial is run towards the end of /etc/rc, not right after the >npx probe? Output interrupts must be broken. They may have always been broken, but 2.1.5 polls for lost ones every 10 msec, so output goes out at at least 100 cps, while 2.2 polls every second, so output goes out at at least 1 cps. You can restore the 2.1.5 polling frequency by setting the flag to recover sooner from lost output interrupts, or by not using interrupts at all. See the sio man page. Bruce