Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
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
Date: Wed, 27 Jul 2005 16:52:29 -0400
From: Jason Tishler <jason AT tishler DOT net>
To: cygwin AT cygwin DOT com
Subject: Re: rebaseall failure?
Message-ID: <20050727205229.GA3876@tishler.net>
Mail-Followup-To: cygwin AT cygwin DOT com
References: <7151719E-2B2E-4A26-8D59-91E2500B8045 AT mac DOT com> <20050727164719 DOT GA1816 AT tishler DOT net> <5782337 DOT 1122488184052 DOT JavaMail DOT dgou AT mac DOT com> <20050727182912 DOT GB8798 AT trixie DOT casa DOT cgf DOT cx> <7429953 DOT 1122490121496 DOT JavaMail DOT dgou AT mac DOT com> <20050727190657 DOT GA2272 AT tishler DOT net> <10309717 DOT 1122492940641 DOT JavaMail DOT dgou AT mac DOT com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <10309717.1122492940641.JavaMail.dgou@mac.com>
User-Agent: Mutt/1.4.1i
X-IsSubscribed: yes
Doug,
On Wed, Jul 27, 2005 at 03:35:40PM -0400, Doug Philips wrote:
> On Wednesday, July 27, 2005, at 03:05PM, Jason Tishler indited:
> >If sh.exe is really bash.exe, then this is why cygiconv-2.dll is in
> >use. The question is why is sh.exe running in the first place?
>
> Indeed. And why does putting that dll at the end of the list make any
> difference?
AFAICT, it doesn't. See below...
On Tue, Jul 26, 2005 at 07:26:17PM -0400, Douglas Philips wrote:
> I copied rebaseall and had my copy just save the zcat filtered output
> into /tmp/rebase.lst .
> I then moved /usr/bin/cygiconv-2.dll to the very end of the list and
> ran:
> $ rebase -d -b 0x70000000 -o 0x10000 -T /tmp/rebase.lst
The above worked because you invoked rebase directly instead of
rebaseall, not because you changed the order of the DLLs rebased.
> My next line of inquiry resulted in putting a 'sleep' in just before
> the rebase invocation. Huh, that fixed it. Weird.
The sleep gave the sh.exe process time to exit. You "fixed" it but in
an open-loop way that is still prone to failure.
> Then I went looking, and what happens just before rebase is called?
> zcat and egrep are called.
> zcat is a link to gzip which is a .exe file.
> egrep, however, is a shell script
> $ cd /usr/bin
> $ cat egrep
> #!/bin/sh
> exec grep -E ${1+"$@"}
> $
The above is causing the problem. Why is egrep a shell script on your
system? It should be a symlink:
$ ls -l /usr/bin/egrep
lrwxrwxrwx 1 jtishler Users 4 Mar 29 2004 /usr/bin/egrep -> grep
Either restored the symlink or change the trigger line to:
#!/bin/ash
> So apparently the shell isn't shutting down fast enough before rebase
> is invoked.
This seems to be the case.
> Of course none of this fixes my real problem with why Python is unable
> to get a lock...
I was able to reproduce your problem. So far, I found from stracing the
following:
7743 884473 [main] python2.4 3916 __set_errno: int semaphore::_trywait():1761 val 11
Unfortunately, I won't have the time to debug further for a while. And,
to be honest, I'm lacking the motivation too...
Jason
--
PGP/GPG Key: http://www.tishler.net/jason/pubkey.asc or key servers
Fingerprint: 7A73 1405 7F2B E669 C19D 8784 1AFD E4CC ECF4 8EF6
--
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/