delorie.com/archives/browse.cgi   search  
Mail Archives: cygwin/2002/05/05/16:54:49

Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com
Message-Id: <5.1.0.14.2.20020505133943.00b04720@pop3.cris.com>
X-Sender: rrschulz AT pop3 DOT cris DOT com
Date: Sun, 05 May 2002 13:54:37 -0700
To: cygwin AT cygwin DOT com
From: Randall R Schulz <rrschulz AT cris DOT com>
Subject: RE: New snapshot with significant new functionality
In-Reply-To: <Pine.LNX.4.44.0205051611330.9796-100000@rd.incert.com>
References: <FC169E059D1A0442A04C40F86D9BA7600C5F9D AT itdomain003 DOT itdomain DOT net DOT au>
Mime-Version: 1.0

Chris, Chris, Chris,

Here are a couple of alternate ideas:

1) Require a mount to access /proc (or just /proc/registry). Use a 
mount-time option to enable writability of the registry portion of /proc 
(an option to the mount command and a bit in the __flags argument to the 
mount system call).

2) Use the writability of /proc/registry. This has the advantage of 
providing some selectivity w.r.t. which user attempts the registry update.


Randall Schulz
Mountain View, CA USA



At 13:14 2002-05-05, you wrote:
>On Fri, 3 May 2002, Robert Collins wrote:
> > Two things: if /proc/registry isn't writable, cating 1 to
> > /proc/registry/.writeable won't work - without special case code. I'd
> > suggest /proc/sysopts/fs/registry/writeable.
> >
> > Two, why not have two options:
> > writeable
> > nextwrite
> >
> > one is persistent (until all cygwin processes end). The other is for a 
> single transaction.
>
>The "single transaction" mode seems like it might be race-prone, if two 
>processes (or two threads of the same process) are both accessing the 
>registry.  Perhaps it would be cleaner to allow a process to open a 
>registry file for write, but then require some fd-specific action (e.g. an 
>ioctl, or perhaps writing the fd number to a /proc/sysopts file) to enable 
>it to *really* work for write.
>
>                 Chris Metcalf -- InCert Software -- 1 (617) 621 8080
>                 metcalf AT incert DOT com -- http://www.incert.com/~metcalf\


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019