From owner-freebsd-questions Tue Jun 8 9:49:33 1999 Delivered-To: freebsd-questions@freebsd.org Received: from netdoor.com (netdoor.com [208.137.128.6]) by hub.freebsd.org (Postfix) with ESMTP id C6FA5151F6 for ; Tue, 8 Jun 1999 09:49:28 -0700 (PDT) (envelope-from dougj@netdoor.com) Received: from [208.137.128.7] (dougj@[208.137.128.7]) by netdoor.com (8.9.3/8.9.3) with ESMTP id LAA29057; Tue, 8 Jun 1999 11:48:53 -0500 (CDT) Date: Tue, 8 Jun 1999 11:48:52 -0500 (CDT) From: Doug Jennings To: "James R. Shrenk" Cc: Mark Ovens , freebsd-questions@FreeBSD.ORG Subject: Re: silo overflow errors In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I am now positive that it is not the modem's fault. I installed Win 98 and the modem is performing quite nicely -- even downloading at 3 k/sec at times. When originally trying to configure and dial out with user ppp, I noticed that my CPU usage was skyrocketing to 90% (and I have a PII 450, so this concerned me to say the least) -- but this only seemed to happen during a failed dial attempt. I am fairly new at BSD, so it took me a day or two to get my modem and ppp.conf playing along nicely. I will check and see if the CPU usage is still as high as it was then. Do you know where I can get some info on how to resolve the problem if it is an IRQ issue? You guys have been really helpful, and I thank you for it. More later, I am sure. -Doug On Tue, 8 Jun 1999, James R. Shrenk wrote: > looks good so far. I took a visit around some man pages and the web and > it seems unlikely that it's actually the modem that would be causing the > problem. Check the output of top during your ppp sessions. In > particular, what is your CPU usage during that time. The data that's > buffered in the UART seems to be dropping in this case. The slow is then > probably a result of packets being retransmitted to account for the > errors. The man page for 3.2-release says that these silo overflows are > the result of a problem with the interrupt handler. Although I'm sketchy > on this, your serial port may not be getting enough attention from the CPU > so it may be worth checking out. > > James > > On Mon, 7 Jun 1999, Mark Ovens wrote: > > > On Mon, Jun 07, 1999 at 05:17:53PM -0500, Doug Jennings wrote: > > > When I get the output of dmesg, it contains the following lines: > > > > > > sio0 at 0x3f8-0x3ff irq 4 flags 0x10 on isa > > > sio0: type 16550A > > > > > > I knew it had to be a 16550A UART because the computer is brand new with all > > > the appropriate bells / whistles / et cetera. Argh. I suppose I will see > > > if I can slow down the modem now. > > > > > > One more question...do you think the problem is happening inside the modem > > > itself? Or could it be caused by some kind of interference in my serial > > > cable? > > > > It would be worth checking the shield on your cable, especially if > > it runs parallel to the mains lead. I've seen silo overflow errors > > generated at the rate of >50 per second on a Sun with serial lines > > to terminals running through a factory close to arc welders :-) > > > > > I am obviously grasping here, but I just can't figure out why I am > > > having silo errors with such a supposedly good modem (external courier > > > v.everything 28800). > > > > > > >What is your dmesg output? It looks like you might have a older UART > > > >(8250 maybe?) If that's the case, you might try slowing the modem down to > > > >14400 and see if that doesn't help. > > > > > > > > > > > >On Mon, 7 Jun 1999, Doug Jennings wrote: > > > > > > > >> > > > >> After finally getting user ppp to work, I keep running into the > > > >> following problem. Once I establish a ppp connection and I try to > > > >> send/receive any data whatsoever (for example, opening a telnet > > > session), I > > > >> get the message: > > > >> > > > >> /kernel: sio0: 1 more silo overflow (total 1) > > > >> > > > >> The 'total' goes up each time I send/receive any data (for example, if I > > > >> issue a "ls" command in the telnet session and have to receive the output > > > of > > > >> that). It slows down my 28800 modem to an absolute crawl. > > > >> > > > >> Any ideas on a fix for this? I am running a fresh installation of FreeBSD > > > >> 3.1 and I am using an external USR v.everything courier modem. I know > > > that > > > >> it has to do with some buffering issue -- I just don't know how to fix > > > it. > > > >> > > > >> Any ideas would be appreciated. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message