*BSD News Article 17553


Return to BSD News archive

Newsgroups: comp.os.386bsd.bugs
Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!osuunx.ucc.okstate.edu!moe.ksu.ksu.edu!ux1.cso.uiuc.edu!howland.reston.ans.net!wupost!uunet!mcsun!sun4nl!eur.nl!pk
From: pk@cs.few.eur.nl (Paul Kranenburg)
Subject: Re: ioctls and core dumps
Message-ID: <1993Jun26.074010.5689@cs.few.eur.nl>
Sender: news@cs.few.eur.nl
Reply-To: pk@cs.few.eur.nl
Organization: Erasmus University Rotterdam
References: <1993Jun22.173715.6498@fac.com> <20cljmINNa7a@fstgds01.tu-graz.ac.at>
Date: Sat, 26 Jun 1993 07:40:10 GMT
Lines: 17

In <20cljmINNa7a@fstgds01.tu-graz.ac.at> chmr@edvz.tu-graz.ac.at (Christoph Robitschko) writes:

[...]
>VM system does with changes, since there are not supposed to be any.
>If you change the protections with mprotect, the VM system does the
>right thing.

Not quite. Currently, upgrading permissions on a mmap'ed region of memory
from, say, read-only to read-write, is not handled correctly by the
VM system. As I've pointed out earlier, this can lead to a kernel deadlock
or, in case the mmap'ed file resides on an NFS filesystem, to a panic due
to IO without a process context or to silent failure of the write back
because of misplaced credentials.

-pk