From owner-freebsd-hackers Sat Feb 24 06:07:21 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id GAA04152 for hackers-outgoing; Sat, 24 Feb 1996 06:07:21 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id GAA04103 for ; Sat, 24 Feb 1996 06:07:12 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id BAA06181; Sun, 25 Feb 1996 01:03:35 +1100 Date: Sun, 25 Feb 1996 01:03:35 +1100 From: Bruce Evans Message-Id: <199602241403.BAA06181@godzilla.zeta.org.au> To: bde@zeta.org.au, msmith@atrad.adelaide.edu.au Subject: Re: 230.400 bps on serial port Cc: hackers@FreeBSD.org, petzi@zit.th-darmstadt.de Sender: owner-hackers@FreeBSD.org Precedence: bulk >> - IDE disks have no effect on serial communications under FreeBSD. >> (Bus-hogging DMA controllers may have an adverse affect if they steal >> too many cycles.) >Hmm, that means splbio() is lower than spltty(), correct? That's a >genuine misunderstanding on my part. They're independent, but this isn't very relevant. spltty() only blocks tty software interrupts. I forgot that IDE disks can interfere with serial communications if they prevent the tty software interrupt handlers from running for more than about 20 msec - the low level buffer space runs out. If this happens then you may have more things to worry about - timeout routines would also be delayed for 20 msec, and it is a short step from there to the system spending all its time in the kernel. Bruce