From owner-freebsd-bugs  Mon May  8 10:59:06 1995
Return-Path: bugs-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id KAA06197
          for bugs-outgoing; Mon, 8 May 1995 10:59:06 -0700
Received: from Root.COM (implode.Root.COM [198.145.90.1])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id KAA06191
          for <freebsd-bugs@FreeBSD.org>; Mon, 8 May 1995 10:59:04 -0700
Received: (dyson@localhost) by Root.COM (8.6.8/8.6.5) id LAA00848; Mon, 8 May 1995 11:01:59 -0700
Date: Mon, 8 May 1995 11:01:59 -0700
From: John Dyson <dyson@Root.COM>
Message-Id: <199505081801.LAA00848@Root.COM>
To: freebsd-bugs@FreeBSD.org, sung!imdave@iwcs.att.com
Subject: Re: Process hangs sleeping in ufslk2
Sender: bugs-owner@FreeBSD.org
Precedence: bulk

When installing pdksh on the April snapshot, you are likely to run
into this problem.  It comes about because of a bug in the
merged VM/Buffer cache code.  (Specifically, that version of the
code could not deal with I/O to/from an mmaped region to the file
that the mmaped region represents.)

Something like this:

fd = open("evilfile", O_RDWR);

addr = mmap(...,fd,...)

write(fd, addr, size);  -- you hang here.

This problem has been fixed in the -current and in the 2.0.5 release
coming up.  It is very likely that many of the "mysterious" ufslk2
hangs had been due to this.

Personal note: Dave, I am at Lisle (810-6136...)

John
dyson@root.com