From owner-freebsd-questions@FreeBSD.ORG Tue Oct 14 05:46:08 2003 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id A0E8916A4B3 for ; Tue, 14 Oct 2003 05:46:08 -0700 (PDT) Received: from smtp0.adl1.internode.on.net (smtp0.adl1.internode.on.net [203.16.214.194]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6EA8943FDF for ; Tue, 14 Oct 2003 05:46:07 -0700 (PDT) (envelope-from malcolm.kay@internode.on.net) Received: from beta.home (ppp48-95.lns1.adl2.internode.on.net [150.101.48.95]) h9ECk45A036574; Tue, 14 Oct 2003 22:16:05 +0930 (CST) Content-Type: text/plain; charset="iso-8859-1" From: Malcolm Kay Organization: At home To: jesse@wingnet.net, freebsd-questions@freebsd.org Date: Tue, 14 Oct 2003 22:16:03 +0930 User-Agent: KMail/1.4.3 References: In-Reply-To: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200310142216.03412.malcolm.kay@internode.on.net> Subject: Re: sio3 (COM4) problems on FreeBSD 4.8-RELEASE X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 14 Oct 2003 12:46:08 -0000 On Tue, 14 Oct 2003 01:37, Jesse Guardiani wrote: > Howdy list, > > I'm having trouble with my sio3 (/dev/cuaa3) port on one > of my servers (with an SMP kernel). When I try to `tip com4` > to a cisco switch using this port, I get only partial data, > and this appears in the logs: > > messages:Oct 10 17:18:37 billmax /kernel: sio3 at port 0x2e8-0x2ef irq = 9 on > isa0 messages:Oct 10 17:18:38 billmax /kernel: sio3: type 16550A > messages:Oct 10 17:20:16 billmax /kernel: sio3: 1 more silo overflow (t= otal > 1) messages:Oct 10 17:20:18 billmax /kernel: sio3: 1 more silo overflow > (total 2) messages:Oct 10 17:20:21 billmax /kernel: sio3: 1 more silo > overflow (total 3) messages:Oct 10 17:20:22 billmax /kernel: sio3: 1 mo= re > silo overflow (total 4) messages:Oct 10 17:20:23 billmax /kernel: sio3:= 1 > more silo overflow (total 5) messages:Oct 10 17:20:27 billmax /kernel: > sio3: 1 more silo overflow (total 6) messages:Oct 10 17:20:36 billmax > /kernel: sio3: 1 more silo overflow (total 7) messages:Oct 10 17:20:38 > billmax /kernel: sio3: 1 more silo overflow (total 8) messages:Oct 10 > 17:20:45 billmax /kernel: sio3: 1 more silo overflow (total 9) messages= :Oct > 10 17:20:48 billmax /kernel: sio3: 1 more silo overflow (total 10) > messages:Oct 10 17:21:04 billmax /kernel: sio3: 1 more silo overflow (t= otal > 11) messages:Oct 10 17:22:15 billmax /kernel: sio3: 1 more silo overflo= w > (total 12) messages:Oct 10 17:40:07 billmax /kernel: sio3: 1 more silo > overflow (total 13) messages:Oct 10 17:46:29 billmax /kernel: sio3: 1 m= ore > silo overflow (total 14) > > > If I move the console cable from com4 to com1 on the server and keep th= e > console cable in the same switch, then `tip com1`, it works fine. It's = just > the sio3 port that I'm having trouble with. Note that this machine only= has > sio1 and sio3. > > Here's my sio kernel config: > > # Serial (COM) ports > device sio0 at isa? port IO_COM1 flags 0x10 irq 4 > device sio1 at isa? port IO_COM2 irq 3 > device sio2 at isa? disable port IO_COM3 irq 5 > device sio3 at isa? port IO_COM4 flags 0x10 irq 9 > > Also, I've tried it with AND without the 0x10 flag (which I know just > allows the com device to be used as a serial console). AND I've tried i= t > with the "do not use FIFO" flag (0x00002) on COM4, like this: > > # Serial (COM) ports > device sio0 at isa? port IO_COM1 flags 0x10 irq 4 > device sio1 at isa? disable port IO_COM2 irq 3 > device sio2 at isa? disable port IO_COM3 irq 5 > device sio3 at isa? port IO_COM4 flags 0x00002 irq 9 > > > > I get the same error in the log either way. Port speed is set to 9600 i= n > all tests. I get the same problem under conserver, so it's not directly > related to the `tip` command, but the device. > > Is there some magic voodoo flag that I need to pass to this port to get= it > working? > Dredging back through memory I seem to recall that irq 9 was somewhat problematic (not specifically a FreeBSD problem) being a remapping of irq= 2 which *sometimes* worked. If possible it might be worth trying to rewire the interrupt to irq 3 or = irq=20 5. But then difficulties of the past with irq 2/9 may have been solved in modern hardware. On the other hand it seems to me that flakey interrupts could give rise to the symptoms you are experiencing. Malcolm=20