X-Recipient: archive-cygwin AT delorie DOT com
X-Spam-Check-By: sourceware.org
From: "Dave Korn" <dave DOT korn AT artimi DOT com>
To: <cygwin AT cygwin DOT com>
References: <081020144229 DOT ZM1710 AT adobe DOT com> <20081023135451 DOT GB21067 AT calimero DOT vinschen DOT de> <20081023140902 DOT GR9289 AT calimero DOT vinschen DOT de> <010e01c93525$ab77b160$9601a8c0 AT CAM DOT ARTIMI DOT COM> <20081023162033 DOT GS9289 AT calimero DOT vinschen DOT de>
Subject: RE: cygwin bash crashes on Win Serv 2008
Date: Thu, 23 Oct 2008 17:52:10 +0100
Message-ID: <011401c9352f$b0e25dd0$9601a8c0@CAM.ARTIMI.COM>
MIME-Version: 1.0
Content-Type: text/plain; 	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
In-Reply-To: <20081023162033.GS9289@calimero.vinschen.de>
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Precedence: bulk
List-Id: <cygwin.cygwin.com>
List-Unsubscribe: <mailto:cygwin-unsubscribe-archive-cygwin=delorie DOT com AT cygwin DOT 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

Corinna Vinschen wrote on 23 October 2008 17:21:

>>   Only thing I can think of is "Not if %ss has been mucked around with it
>> isn't".
> 
> Yeah, I heard about that.  But what is %ss doing in Windows 

  Same as usual.  Pointing to the stack segment.  It just /happens/ that the
SS is a full 32-bit flat mapping of the same virtual memory space as %cs, %ds,
etc,. but it doesn't (in theory) have to be.

> and why should it be messed up with TS?!?  

  Extra security measures turned on by default on multi-user servers that are
turned off by default usually?

> And why are only Cygwin processes affected, and then only some?

  Just lucky, I guess...


>>   I'd use windbg on this, take a look at the exception record and
>> selectors and stuff.
> 
> The exception record was a good hint (I don't know what you mean by
> "selectors", sorry).  

  See the "dg" command, e.g. "dg %cs", "dg %ds", "dg %ss".

> Unfortunately it puzzles me even more:
> 
>     ExceptionAddress: 00419d97 (image00400000+0x00019d97)
>        ExceptionCode: c0000005 (Access violation)
>       ExceptionFlags: 00000000
>     NumberParameters: 2
>        Parameter[0]: 00000008
>        Parameter[1]: 00419d97
>     Attempt to execute non-executable address 00419d97
> 
> Huh?  Why should this address (this application function) be
> "non-executable", while it's executable when TS is not installed?

  DEP?  ASLR?  SafeSEH?  As well as "dg" there are some other commands in
windbg that'll show you memory types and attributes.

> Could this have something to do with the executbale header gcc creates?

  Dunno - which executable header?  Seems unlikely since we're in a completely
different memory page and well beyond the header area into the .text segment.

  (Actually, are we in the .text segment, or is there a thunk of some kind in
.rdata?  And is the difference perhaps related to the use-or-not, or the
need-for-or-not, of ld's --{en,dis}able-runtime-pseudo-reloc options?)

    cheers,
      DaveK
-- 
Can't think of a witty .sigline today....


--
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/