From owner-freebsd-current Thu Dec 31 21:30:04 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA03292 for freebsd-current-outgoing; Thu, 31 Dec 1998 21:30:04 -0800 (PST) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from apollo.backplane.com (apollo.backplane.com [209.157.86.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id VAA03279 for ; Thu, 31 Dec 1998 21:30:01 -0800 (PST) (envelope-from dillon@apollo.backplane.com) Received: (from dillon@localhost) by apollo.backplane.com (8.9.1/8.9.1) id VAA09794; Thu, 31 Dec 1998 21:29:34 -0800 (PST) (envelope-from dillon) Date: Thu, 31 Dec 1998 21:29:34 -0800 (PST) From: Matthew Dillon Message-Id: <199901010529.VAA09794@apollo.backplane.com> To: Tony Overfield Cc: current@FreeBSD.ORG Subject: Re: processes hanging in "inode" state? Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG :Lately, I have been playing with -current on a test system. When :I do a "make -j 8 world", everything always works fine with small :to moderate amounts of swap usage. When I do a "make -j 32 world", :as expected, I encounter a greatly increased amount of swap usage. :However, it doesn't ever finish. After a long time, everything :simply stops and it cannot start any new programs. If I happen to :have top already running, I can switch to it and see numerous :... : :I have no reason to suspect hardware trouble and I think I have :plenty of swap space for this to complete, if only it didn't hang. : :What should I do to find (or help somebody find) the problem? : :P5/MMX 233MHz, 430FX/PIIX3, 64MB EDO DRAM, 2940U, Quantum/DEC DSP3210. : :- Tony It's probably a memory deadlock somewhere. If you have DDB configured, you can control-alt-escape into the debugger and do a 'ps' to see what various processes are waiting on. That kind of dump would be very helpful. Most of the memory related deadlock issues will be resolved after the 3.0.1 release and the tree split, assuming David Greenman and I can come to agreement (which I think is likely). I've been working on a nearly complete replacement for the swap pager and along with that comes a whole slew of fixes for various low-memory deadlock conditions. I'm using a diskless bootp workstation with MFS filesystems and NFS swap as a testbed - a combination that tends to bring out the worse in the kernel :-) -Matt Matthew Dillon Engineering, HiWay Technologies, Inc. & BEST Internet Communications & God knows what else. (Please include original email in any response) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message