[cvsnt] Re: bug in cvs(nt)?

Paul G. pgarceau at comcast.net
Sun Jun 13 21:39:05 BST 2004



On 12 Jun 2004 at 12:00, cvsnt-request at cvsnt.org wrote:

> Date: Sat, 12 Jun 2004 08:06:43 +0200
> From: "Oliver Giesen" <ogware at gmx.net>
> Subject: [cvsnt] Re: bug in cvs(nt)?
> To: cvsnt at cvsnt.org
> Message-ID: <cae3di$dar$1 at paris.nodomain.org>
> Content-Type: text/plain;	charset="iso-8859-1"
> 
> > All three crashed.  Same exception I noted previously.
> 
> OK, but obviously it did *not* crash in at least one situation.

	Hmm...may be some confusion about term "crash".  When I say "crash", I mean that some process aborted 
for some "unknown" or otherwise "undefined" reason.

	In this case, there was a "crash" in every single case -- ie. every single time I initiated the process, cvs.exe 
(cvsnt) had an application error:

	[cvs.exe]
	"The exception unknown software exception (0xc06d007e) ocurred in the application at location  
0x77f1d642." 

	One of the things I do, a form of QA, is to "accept" the "crash", and see what happens if I continue  with (or 
"complete") any given process(es) after a "crash" has ocurred.

	In this case, I hit "ok" when an application error occurs in order to determine what impact that "crash" had on 
the operations of the "processes" in question.

::Processes:

	Occurs when:

	"cvs info" entered at dos box (NT4 cmd.exe, within "cvsnt" folder/directory).

	"cvs info" entered via WinCvs menu selection "Admin|Command line" and hit "ok" 
without "/n" (press "ok" in dialog window).

	"cvs info" entered via WinCvs menu selection "View|Output" window.

	The next thing that occurs:

{
	[cvs.exe]
	"The exception unknown software exception (0xc06d007e) ocurred in the application at location  
0x77f1d642." 
	(options/buttons available:  "ok" or "cancel")
}

--	Hit "ok".
	
	--Yields following WinCvs output on bottom of WinCvs gui to window "opened&docked" via "View|Output" 
menu selection.

Available protocols:

local               (internal)
ext                 ext 2.0.41a
fork                fork 2.0.41a
ntserver            ntserver 2.0.41a
pserver             pserver 2.0.41a
server              server 2.0.41a
sserver             sserver 2.0.41a
ssh                 ssh 2.0.41a

[The output above is reproducible.]

-- hitting "cancel" yields output from my jit debugger, DrMingw.

::End Processes Descriptions.

Using WinCvs menu selection "Remote|Checkout module"; __does not yield same results__.

> Where else
> did you get that example output you quoted earlier?

> Or are you saying it
> crashed and nevertheless produced the correct output?

	I do not know how "correct output" is defined as far as WinCVS (on return from cvs.exe) is concerned after 
"cvs info" has been requested via the processes noted above.

	I have assumed that the prompt re: "Available Protocols" (as noted by WinCvs.exe), is an indication of a 
problem with the abortion ("crash") of the cvs.exe application that is caused by the cvs.exe application error noted 
above.

> 
> Tony, have you got any idea how that "missing DBGHLP.DLL" problem might be
> resolved so that Paul could at least produce a meaningful crashdump to
> send to you?

	Please?

	Thank you.

		Paul G.



More information about the cvsnt mailing list