From owner-freebsd-hackers Fri Apr 23 15:26:18 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from dingo.cdrom.com (castles539.castles.com [208.214.165.103]) by hub.freebsd.org (Postfix) with ESMTP id 200B31557E for ; Fri, 23 Apr 1999 15:25:10 -0700 (PDT) (envelope-from mike@dingo.cdrom.com) Received: from dingo.cdrom.com (localhost [127.0.0.1]) by dingo.cdrom.com (8.9.3/8.8.8) with ESMTP id TAA00635; Thu, 22 Apr 1999 19:15:02 -0700 (PDT) (envelope-from mike@dingo.cdrom.com) Message-Id: <199904230215.TAA00635@dingo.cdrom.com> X-Mailer: exmh version 2.0.2 2/24/98 To: Luoqi Chen Cc: dick@tar.com, jplevyak@inktomi.com, hackers@FreeBSD.ORG Subject: Re: flock + kernel threads bug In-reply-to: Your message of "Tue, 20 Apr 1999 17:19:57 EDT." <199904202119.RAA17889@lor.watermarkgroup.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 22 Apr 1999 19:15:02 -0700 From: Mike Smith Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > No, this is not an issue of separate task and thread structures. You may > have the same problem for two separate processes that share file descriptor > table, and john's fix would not work for this case. The correct solution > to this problem (in my opinion) is to clear all POSIX locks upon closure > of a file descriptor, which requires we keep in the file descriptor table > a list of procs that reference it. Hmm. We seem to need some better process/resource affinity tracking. I've just recently been looking at some code that wants for a callout on process exit; rather than adding yet more code to the process exit path it'd be nice to have a callout list for process events (fork, exit, etc.) More to think about I guess. -- \\ Sometimes you're ahead, \\ Mike Smith \\ sometimes you're behind. \\ mike@smith.net.au \\ The race is long, and in the \\ msmith@freebsd.org \\ end it's only with yourself. \\ msmith@cdrom.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message