From fnaumann@mail.cs.uni-magdeburg.de Sun May 30 20:58:40 2004 Date: Sun, 30 May 2004 14:55:21 -0400 (EDT) From: Jim DeClercq To: mint_list , support@panix.com Subject: [MiNT] Possible pppd based my-system crashes Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Virus-Scanned: by amavisd-new-20030616-p7 (Debian) at fishpool.fi Delivered-To: mint@fishpool.com Delivered-To: mint@lists.fishpool.fi X-ecartis-version: Ecartis v1.0.0 Sender: mint-bounce@lists.fishpool.fi Errors-to: mint-bounce@lists.fishpool.fi X-original-sender: jimd@panix.com Precedence: bulk List-help: List-unsubscribe: List-ID: X-List-ID: X-Milter: ClamAV 0.70/0.70jjel X-Milter: milter-regex 1.5jel X-Milter: ClamAV 0.70/0.70jjel X-Milter: milter-regex 1.5jel I get quite a few crash and reboot errors when doing anything other than reading mail or newsgroups. That is when I am typing e-mail or a reply to e-mail over a ppp link. These show up as aessys errors, segmentation errors, bash errors, toswin2 errors, and sld errors. Most of the error reports have something to do with a tcp_send seg outside win. I do not know what is complaining about who, in this context, or whether an outgoing error or incoming error is involved. Pppd somewhere is complaining about something. Most of them happen when I am typing fast, my normal speed, but I tried typing slowly, and that did not help either. Anything coming from my keyboard, TT030, TTM 195, MiNT 1.15.12, will eventually, and quite quickly, cause a system crash, working on e-mail, in pico and pine, over a ppp link. There are three areas of commonality possible. The first is MiNT's keyboard buffer, on the assumption it has one, being either too small or too slow, or too fast. The second is whatever buffer pppd sets up, and how it interacts with the keyboard buffer. The third is netBSD's pppd. If anyone is looking through the MiNT source code for either of those, and finds a buffer being set up, it would be a boon if they would make it a little bigger, and my mean-time-to-failure on typing something on a remote host might go up a little. If anyone could suggest another place where this problem might be hiding, that too would be appreciated. At this point, I do not know where to look. A further complication is that I have no crashes, repeat, no crashes, when replying to a UseNet newsgroup posting. It might even be at the other end. However, what I use for a newsreader is different than they provide. I have tin aliased to tinx. This is also somewhat of a new problem, and my remote host has upgraded versions of netBSD recently. Somebody somewhere has a buffer overrun problem, I think. If anyone anywhere could suggest where it might be, I would be greatful. This is being typed off line, and sent by FTP to my remote host, so such problems will not happen. Jim -- /"\ Jim DeClercq--jimd@panix.com--Sylvania, Ohio, USA \ / ASCII ribbon campaign | I'm a .signature virus! | X against HTML mail | Copy me into your ~/.signature| / \ and postings | to help me spread! | .