From owner-cvs-sys Sun Jul 2 03:19:35 1995 Return-Path: cvs-sys-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id DAA24429 for cvs-sys-outgoing; Sun, 2 Jul 1995 03:19:35 -0700 Received: from bitcom (sovcom.kiae.su [144.206.136.1]) by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id DAA24422 ; Sun, 2 Jul 1995 03:19:21 -0700 Received: by bitcom id AA07524 (5.65.kiae-1 ); Sun, 2 Jul 1995 13:12:18 +0300 Received: by sovcom.KIAE.su (UUMAIL/2.0); Sun, 2 Jul 95 13:12:18 +0300 Received: by astral.msk.su (UUPC/@ v6.14g, 06Jun95) id AA04533; Sun, 2 Jul 1995 12:34:57 +0400 (MSD) To: CVS-commiters@freefall.cdrom.com, Bruce Evans , cvs-sys@freefall.cdrom.com References: <199506300206.MAA21212@godzilla.zeta.org.au> In-Reply-To: <199506300206.MAA21212@godzilla.zeta.org.au>; from Bruce Evans at Fri, 30 Jun 1995 12:06:11 +1000 Message-Id: Organization: Ha-olahm Yetzirah X-Mailer: Mail/@ [v2.38 MSDOS] Date: Sun, 2 Jul 1995 12:34:56 +0400 (MSD) From: =?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?= aka "Andrey A. Chernov, Black Mage" X-Class: Fast Subject: Re: cvs commit: src/sys/i386/isa sio.c Mime-Version: 1.0 Lines: 38 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Length: 1712 Sender: cvs-sys-owner@freebsd.org Precedence: bulk In message <199506300206.MAA21212@godzilla.zeta.org.au> Bruce Evans writes: >> Modified: sys/i386/isa sio.c >> Log: >> Fight with hanging modems continued: >> return EIO after t_timeout expired instead infinite looping in "siotx" >> in comparam, consuming CPU time. >Why does this loop occur? It can't happen with working 8250..16550 >chips? It might happen if the UART incompatibly implements ctsflow >in hardware. Otherwise, hangs here are quite different from the >ctsflow-related hangs solved elsewhere using t_timeout. Bruce, I don't know, *why* it occurs, I am not hardware expert. I only know that it *occurs* sometimes and my goal is to make modem pool which can work without human eye when possible. This thing cures by dropping DTR, EIO here usualy cause close and DTR drop. I am not think, that it is CTS hang, symptom are: modem TX light is on forever (sio waits, when TX done forever too). Joerg f.e. send me "thanx" for this fix, he have this situation too (it means that it isn't only my hardware). After few years of dealing with modem pools I know, that *all* hardware are broken and good hardware simple doesn't exists. It means, that sio/tty driver must *not* have infinite loops and have reasonable timeout for each potential long-time operation. I understand, that even now not all such places handled with timeouts, but I prefer to fix only *real* bugs, not potential ones. -- Andrey A. Chernov : And I rest so composedly, /Now, in my bed, ache@astral.msk.su : That any beholder /Might fancy me dead - FidoNet: 2:5020/230.3 : Might start at beholding me, /Thinking me dead. RELCOM Team,FreeBSD Team : E.A.Poe From "For Annie" 1849