Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Message-ID: <48270-2200210431175433137@M2W094.mail2web.com>
Reply-To: lhall AT rfk DOT com
From: "lhall AT pop DOT ma DOT ultranet DOT com" <lhall AT pop DOT ma DOT ultranet DOT com>
To: cbfalconer AT worldnet DOT att DOT net, cygwin AT cygwin DOT com
Subject: Re: gdb hangs on a 486
Date: Thu, 31 Oct 2002 12:54:33 -0500
MIME-Version: 1.0
Content-type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit


>From: CBFalconer cbfalconer AT yahoo DOT com
>Date: Thu, 31 Oct 2002 11:20:45 -0500
>To: cygwin AT cygwin DOT com
>Subject: Re: gdb hangs on a 486
>
>
>"Larry Hall (RFK Partners, Inc)" wrote:
>> 
>> At 11:13 PM 10/30/2002, CBFalconer wrote:
>> >I have been trying out gdb in Cygwin, and found it to hang and/or
>> >crash under W98, running on a 486.  The output of gdb --version
>> >is:
>> >
>>... snip ...
>> > > This GDB was configured as "i686-pc-cygwin".
>> >                               ^^^^^^^^^^^^^^
>> >This appears unwarranted.  I would have assumed gdb would test and
>> >adapt itself to the processor on which it is running.
>> 
>> At this point, I think most (all?) Cygwin packages are configured
>> like this. Whether or not that's true, it's not unwarranted. There's 
>> good reason to make use of the newer architectures' capabilities.
>
>I can easily believe that.  It seems very poor practice to make
>these assumptions without checking them somewhere and generating a
>warning.  Such things can go in initialization or loading code.


Sure.  I doubt many programs actually do architecture checks as part
of their startup though.  If you're arguing that this should be the 
case, I won't contest it.  In a perfect world, this would probably
be the case.  It's a low enough pain-threshold for the majority that 
the lack of this check is not noticed.  But that's a separate issue 
from your point.


>... snip ...
>> 
>> gdb -nw
>> 
>
>Are you saying that the problem is limited to the GUI interface? 
>Is this known, or just a guess?


Actually, no I'm not saying that.  However, you mentioned that gdb 
came up in the GUI version, which I read as preference not to have the
GUI.  That's really all I was pointing out.  It's possible that the 
command line version would cause less windowing/mouse problems though.
I don't expect that the command line version targets i486 while the GUI
targets something later however.


>> 
>> Sounds like you may want to get the source, reconfigure, and build
>> your own version targeting i386 or i486.
>
>A non-trivial job, especially if the very tools are suspect.


A potentially non-trivial job, yes, depending on your skills and 
experience building packages.  I'm not sure what you're referring to
by "the very tools are suspect".  These tools have been around for a
long time.  They worked when these architectures were the default
configuration.  It shouldn't be too hard to get them working on 
those targets now.  If you're referring to the fact that the tools 
don't check if the run-time environment matches the configuration 
environment on start-up, I think labeling the tools as "suspect" for 
this oversight is a little extreme.  But I may be missing your meaning.

Larry

--------------------------------------------------------------------
mail2web - Check your email from the web at
http://mail2web.com/ .


--
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/