From owner-freebsd-current Thu Sep 10 03:40:43 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id DAA05477 for freebsd-current-outgoing; Thu, 10 Sep 1998 03:40:43 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from pcnet1.pcnet.com (pcnet1.pcnet.com [204.213.232.3]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id DAA05472 for ; Thu, 10 Sep 1998 03:40:42 -0700 (PDT) (envelope-from eischen@vigrid.com) Received: (from eischen@localhost) by pcnet1.pcnet.com (8.8.7/PCNet) id GAA03824; Thu, 10 Sep 1998 06:40:39 -0400 (EDT) Date: Thu, 10 Sep 1998 06:40:39 -0400 (EDT) From: Daniel Eischen Message-Id: <199809101040.GAA03824@pcnet1.pcnet.com> To: eischen@vigrid.com, tlambert@primenet.com Subject: Re: Thread Problems Cc: freebsd-current@FreeBSD.ORG, info@highwind.com Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > > I tried to repeat the reported problem, but couldn't. I had > > one thread listening for client connections, a worker thread > > (printing a message once a second), and main. The thread > > listening for client connections would create a thread to > > handle the connection and go back to accepting more connections. > > I couldn't make it fail - the worker thread was always spewing > > out messages once a second. I even tried using fork to > > process each client request that came in - that worked also. > > We really need to get more information. > > Effectively, we have the select coming true before the accept call, > followed by the accept call blocking, even though the select was > true. > > In theory, this can't happen, even with a blocking fd. Actually, I think this can happen. We were able to reproduce this (under HP-UX, didn't try FreeBSD) but I forget the exact circumstances. I think it had something to do when the remote client attempted to make a connection but closed the socket before the connection was complete. I'll have to see if I can find it in my notes - it was 3 or 4 years ago... > This has got to be a code specific problem. Perhaps they made the > mistake of the LDAP implementors about number of descriptors, or > the meaning of "preemptive", and are engaged in a buzz-loop. 8-(. > > I just assumed that the person was running -current, and was seeing > what he thought he was seeing; on the other hand... I think he was running current - don't know how recent it was though. > It would be interesting to know what *exact* version of FreeBSD he > was running, as well. > > Also, if this is a Linux program running under emulation, it should > not be expected to work... No, I remember it being code they were porting to FreeBSD. Dan Eischen eischen@vigrid.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message