X-Spam-Check-By: sourceware.org
Message-ID: <459413FE.8050500@tlinx.org>
Date: Thu, 28 Dec 2006 10:59:10 -0800
From: Linda Walsh <cygwin AT tlinx DOT org>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: cygwin AT cygwin DOT com
Subject: Re: Updated cygwin dll's cause unnecessary reboot on NT
References: <458EE598 DOT 3010404 AT aim DOT com> <458F31B1 DOT 6050804 AT byu DOT net>  <458F81CC DOT 3090500 AT tlinx DOT org> <458FDC4E DOT 9040505 AT cygwin DOT com> <458FEC2E DOT 70705 AT tlinx DOT org>  <Pine DOT GSO DOT 4 DOT 63 DOT 0612251346240 DOT 27982 AT access1 DOT cims DOT nyu DOT edu> <45902BC4 DOT 50803 AT tlinx DOT org>  <4590BD4E DOT 5020905 AT cygwin DOT com> <45910426 DOT 9030603 AT tlinx DOT org> <459133CB DOT 3080102 AT ukf DOT net>  <45916391 DOT 1090906 AT tlinx DOT org> <Pine DOT GSO DOT 4 DOT 63 DOT 0612261313480 DOT 27982 AT access1 DOT cims DOT nyu DOT edu> <45919F7B DOT 8000702 AT tlinx DOT org> <4591FA3C DOT 3020402 AT cygwin DOT com>
In-Reply-To: <4591FA3C.3020402@cygwin.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-IsSubscribed: yes
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Id: <cygwin.cygwin.com>
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/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

Larry Hall (Cygwin) wrote:

> This can be true.  However, to do this effectively, one has to listen and
> learn as well.  Simply restating a point you've made before, if it hasn't
> already been agreed to, does less to convince and more to polarize.  If
> your points aren't agreed upon by the people you're trying to convince,
> it's up to you to take the feedback you're given and figure out how that
> affects your argument.  If you discount the feedback, it doesn't matter
> whether you're on the right or wrong side of the argument.  The fact that
> you don't address the points raised means that you've conceded that 
> there's
> no way to reasonably convince the ones you're lobbying to take up your
> cause.  If that's the case, then further discussion of the point(s) of
> contention is irrelevant and probably even counter-productive.
---
   Thank you for the observation.  I sometimes don't realize how others
may be perceiving the conversation. I tend to over-focus on specific
parts in some conversations and not address issues that seem
tangential or not related.

> Linda, I think we've reach this point here, unless you're willing to try
> to understand why "fork()" is important to Cygwin.
---
  I wasn't aware that anyone thought I was trying to change fork.
I'm not suggesting this. 

>   No one on this list
> is going to be willing to break "fork()" just to make the experience
> of using "setup.exe" a little easier on XP and later O/Ss.
---
   The code for "fork" is part of the cygwin1.dll.  Changes needed to
make setup work are only part of setup.  No changes to
cygwin's internals are needed.

I saw a msg. about deleting runtime libraries from a running
process like "bash". I believe this was meant to show that cygwin
simulates POSIX correctness in fork by manually copying and
reloading DLL's from the pre-forked process to the post-fork process's 
address space. It does this by reloading DLL's from their
original filenames at the time of fork.  This suffers from the
"race-condition" that if the original DLL's change (or in this
case, are updated), then the "post-forked" copy of the program
would not be a POSIX correct copy of the original.

In other words, if someone renames and moves a new DLL into 
place during setup, while a program (ex. bash) is running, 
then, on fork() cygwin may, _incorrectly_, duplicate the 
new address space.   This would cause a technical violation 
of the POSIX requirement that the address space be duplicated.  
Am I misunderstanding this?  

   It comes down to a previously raised point that many (most?)
engineers cannot (or will not) make a trade-off between what is 
technically correct and doing what user's want.  In this case the 
choice is between a) POSIX correctness -- where processes running 
before setup would continue to use old libraries after an update
or b), updating the libraries outside of POSIX as the user is
requesting via setup.

    If POSIX correctness was always the best possible solution
we wouldn't see so many Gnu utilities with non-POSIX
behaviors -- like "bash".  The default is to allow _extended_
functionality unless "POSIX_CORRECT" is set in the environment
(or "--posix" is specified on the command line). If this is
the only issue, the couldn't setup do likewise?

    Am I to understand that this is the only issue preventing
acceptance?

L








--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/