From owner-freebsd-hackers  Wed Apr 21 21:15:26 1999
Delivered-To: freebsd-hackers@freebsd.org
Received: from alcanet.com.au (border.alcanet.com.au [203.62.196.10])
	by hub.freebsd.org (Postfix) with ESMTP id 2861115935
	for <hackers@FreeBSD.ORG>; Wed, 21 Apr 1999 21:15:10 -0700 (PDT)
	(envelope-from peter.jeremy@auss2.alcatel.com.au)
Received: by border.alcanet.com.au id <40349>; Thu, 22 Apr 1999 13:58:45 +1000
Date: Thu, 22 Apr 1999 14:12:30 +1000
From: Peter Jeremy <peter.jeremy@auss2.alcatel.com.au>
Subject: Re: Deadlock on -stable
To: crossd@cs.rpi.edu
Cc: hackers@FreeBSD.ORG
Message-Id: <99Apr22.135845est.40349@border.alcanet.com.au>
Sender: owner-freebsd-hackers@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

"David E. Cross" <crossd@cs.rpi.edu> wrote:
> (I ask since this is a really nasty program, using semaphores,
>shared memory, file locking, mmap()-ed files, file locking, etc.)

It's worthwhile noting that there are a variety of bugs in the
semaphore code.  (I have fixes, but haven't tested them yet).  Most of
these only affect boundary conditions, but there is one case where
semop() can return (with an error) after processing part of the sops
array without rolling back.  None of the bugs should lead to panic's
though.

There was a bug where the semaphore code could tread on innocent
kernel memory, but that was fixed just before 3.1-RELEASE.

Peter


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message