• Rainer Schuetze <r.sagitario gmx.de> Jun 05 2009
• "Vladimir Panteleev" <thecybershadow gmail.com> Jun 05 2009
• Rainer Schuetze <r.sagitario gmx.de> Jun 05 2009
• torhu <no spam.invalid> Jun 08 2009
• Rainer Schuetze <r.sagitario gmx.de> Jun 09 2009
• torhu <no spam.invalid> Jun 11 2009
• Bill Baxter <wbaxter gmail.com> Jun 09 2009
• "Nick Sabalausky" <a a.a> Jun 09 2009
• torhu <no spam.invalid> Jun 11 2009

Rainer Schuetze <r.sagitario gmx.de> writes:

Hi,

cv2pdb is a converter of DMD CodeView debug information to PDB debug 
format to make debugging of D applications that were created with the 
Digital Mars compiler, as seamless as possible in current versions of 
Visual Studio.

Major changes since version 0.1:

     * now works on executables compiled by DMC
       (at least works for debugging DMD)
     * removes .debug section from executable
     * new addin to correctly display D style strings
       and derived object type

cv2pdb is available at

     http://www.dsource.org/projects/cv2pdb

Happy debugging,
Rainer Schuetze

"Vladimir Panteleev" <thecybershadow gmail.com> writes:

On Fri, 05 Jun 2009 10:37:38 +0300, Rainer Schuetze <r.sagitario gmx.de>  
wrote:

 Hi,

 cv2pdb is a converter of DMD CodeView debug information to PDB debug  
 format to make debugging of D applications that were created with the  
 Digital Mars compiler, as seamless as possible in current versions of  
 Visual Studio.

 Major changes since version 0.1:

      * now works on executables compiled by DMC
        (at least works for debugging DMD)
      * removes .debug section from executable
      * new addin to correctly display D style strings
        and derived object type

 cv2pdb is available at

      http://www.dsource.org/projects/cv2pdb

 Happy debugging,
 Rainer Schuetze


Hi, did you notice the two tickets I filed yesterday? I haven't tried 0.3  
yet, but I couldn't get 0.2 to work at all.

-- 
Best regards,
  Vladimir                          mailto:thecybershadow gmail.com

Rainer Schuetze <r.sagitario gmx.de> writes:

Hi,

I hadn't checked the tickets (I still have to set it up correctly to be 
notified, I guess). I could reproduce ticket #2, but not #1.

We should move discussion about bugs there...

Rainer

Vladimir Panteleev wrote:
 On Fri, 05 Jun 2009 10:37:38 +0300, Rainer Schuetze <r.sagitario gmx.de> 
 wrote:
 
 Hi,

 cv2pdb is a converter of DMD CodeView debug information to PDB debug 
 format to make debugging of D applications that were created with the 
 Digital Mars compiler, as seamless as possible in current versions of 
 Visual Studio.


 Hi, did you notice the two tickets I filed yesterday? I haven't tried 
 0.3 yet, but I couldn't get 0.2 to work at all.
 

torhu <no spam.invalid> writes:

On 05.06.2009 09:37, Rainer Schuetze wrote:
 Hi,

 cv2pdb is a converter of DMD CodeView debug information to PDB debug
 format to make debugging of D applications that were created with the
 Digital Mars compiler, as seamless as possible in current versions of
 Visual Studio.


This is great stuff!  Maybe the Windows debugger problem is more or less 
solved with this project.

As someone mostly used to msvc 6, I have to say that debugging in msvc 9 
is incredibly sluggish in comparison.  Most actions in the debugger 
takes several seconds to complete, while in msvc 6 it's more or less 
instantaneous.  Is that also the case when debugging C++?

Rainer Schuetze <r.sagitario gmx.de> writes:

The last time I've been using VC6 is so long ago, I cannot really 
remember... VC7 and VS.NET are maybe a little faster than VS2005/VS2008, 
but to make them stall for seconds is not common neither for C++ nor for D.

The visualizer macros in autoexp.dat are quite complex and can be slow, 
especially if you have associative arrays with a lot of entries. Please 
try it without the autoexp.dat entries.

Rainer

torhu wrote:
 On 05.06.2009 09:37, Rainer Schuetze wrote:
 Hi,

 cv2pdb is a converter of DMD CodeView debug information to PDB debug
 format to make debugging of D applications that were created with the
 Digital Mars compiler, as seamless as possible in current versions of
 Visual Studio.


 This is great stuff!  Maybe the Windows debugger problem is more or less 
 solved with this project.
 
 As someone mostly used to msvc 6, I have to say that debugging in msvc 9 
 is incredibly sluggish in comparison.  Most actions in the debugger 
 takes several seconds to complete, while in msvc 6 it's more or less 
 instantaneous.  Is that also the case when debugging C++?

torhu <no spam.invalid> writes:

On 09.06.2009 20:44, Rainer Schuetze wrote:

 The visualizer macros in autoexp.dat are quite complex and can be slow,
 especially if you have associative arrays with a lot of entries. Please
 try it without the autoexp.dat entries.


Didn't seem to matter much, but then there were no assoc arrays in the 
part of the code I debugged.

Bill Baxter <wbaxter gmail.com> writes:

On Mon, Jun 8, 2009 at 3:43 PM, torhu<no spam.invalid> wrote:
 On 05.06.2009 09:37, Rainer Schuetze wrote:
 Hi,

 cv2pdb is a converter of DMD CodeView debug information to PDB debug
 format to make debugging of D applications that were created with the
 Digital Mars compiler, as seamless as possible in current versions of
 Visual Studio.


 This is great stuff! =A0Maybe the Windows debugger problem is more or les=


 solved with this project.

 As someone mostly used to msvc 6, I have to say that debugging in msvc 9 =


 incredibly sluggish in comparison. =A0Most actions in the debugger takes
 several seconds to complete, while in msvc 6 it's more or less
 instantaneous. =A0Is that also the case when debugging C++?


I had an install of msvc 9 that was so incredibly slow that I just
couldn't bring myself to use it.  But other computers I've used it on
are fine.  There was nothing particularly slow about that PC where it
was running slowly, though.  I think it was just a botched install or
something.  You might try re-installing.

--bb

"Nick Sabalausky" <a a.a> writes:

"Bill Baxter" <wbaxter gmail.com> wrote in message 
news:mailman.242.1244575550.13405.digitalmars-d-announce puremagic.com...
On Mon, Jun 8, 2009 at 3:43 PM, torhu<no spam.invalid> wrote:
 As someone mostly used to msvc 6, I have to say that debugging in msvc 9 
 is
 incredibly sluggish in comparison. Most actions in the debugger takes
 several seconds to complete, while in msvc 6 it's more or less
 instantaneous. Is that also the case when debugging C++?


I had an install of msvc 9 that was so incredibly slow that I just
couldn't bring myself to use it.  But other computers I've used it on
are fine.  There was nothing particularly slow about that PC where it
was running slowly, though.  I think it was just a botched install or
something.  You might try re-installing.


I used to be a huge user of Visual Studio (starting with about version 4 or 
5, IIRC), but I've found that every version of VS starting with the first 
VS.NET has been noticably slower that the previous (With VS6 of course being 
the fastest). Because of this, I've been using VS.NET 2008 Express only when 
I really have to use C#, and Programmer's Notepad 2 for everything else. 

torhu <no spam.invalid> writes:

On 09.06.2009 21:25, Bill Baxter wrote:
 On Mon, Jun 8, 2009 at 3:43 PM, torhu<no spam.invalid>  wrote:
  As someone mostly used to msvc 6, I have to say that debugging in msvc 9 is
  incredibly sluggish in comparison.  Most actions in the debugger takes
  several seconds to complete, while in msvc 6 it's more or less
  instantaneous.  Is that also the case when debugging C++?


 I had an install of msvc 9 that was so incredibly slow that I just
 couldn't bring myself to use it.  But other computers I've used it on
 are fine.  There was nothing particularly slow about that PC where it
 was running slowly, though.  I think it was just a botched install or
 something.  You might try re-installing.

 --bb


I upgraded from msvc 2008 express to 2008 express with service pack 1, 
and it seems to be a bit faster now. Usable, I think.  The installer 
actually uninstalls the old version first, so I don't know if a 
reinstall of the same version would have had the same effect.