Debugging at the next level...

[Timo Kreuzer]

Source level debugging thanks to MSVC and pdb files.
Only the line numbers seem to be a bit off.

[Ged Murphy]

This makes me very _VERY_ happy!!!!

Today is going to be a good day J

Ged.

From: [e-mail] [mailto:[e-mail]] On Behalf Of Timo Kreuzer
Sent: 05 February 2010 04:28
To: ReactOS Development List
Subject: [ros-dev] Debugging at the next level...

Source level debugging thanks to MSVC and pdb files.
Only the line numbers seem to be a bit off.

[Jose Catena]

Great work Timo!!!

I was also using windbg with pdb with my 32 bit mscv kernel version, it really makes debugging so more powerful and easy… PDB support was one of the main reasons I wanted to build with msvc.

I hope to finish very soon a few modifications to the new trap handler so that it can also be compiled with msvc. Currently the traps are working and the interrupts will soon.

Once I have the 32bit kernel compiled with msvc working again, I’ll check your stuff to see how could the 32 and 64 bit builds share the most.

I’m afraid that Sir Richard didn’t understand that his trap handler can’t be compiled with anything but gcc, most probably my fault coz my unpolite way to communicate it.

But I hope that after I have my branch booting and he could check it out, will accept my apologies and reconsider adopting the modifications, coz otherwise the trunk will not be compatible with anything but gcc.

Jose Catena

DIGIWAVES S.L.

[Aleksey Bragin]

On Feb 5, 2010, at 1:00 PM, Jose Catena wrote:
> I’m afraid that Sir Richard didn’t understand that his trap handler  
> can’t be compiled with anything but gcc, most probably my fault coz  
> my unpolite way to communicate it.
>
> But I hope that after I have my branch booting and he could check  
> it out, will accept my apologies and reconsider adopting the  
> modifications, coz otherwise the trunk will not be compatible with  
> anything but gcc.

I will mediate any disputes (hopefully there will be none, of  
course), and I am sure rich wants to have as more portable code as  
possible. I think when trap handlers are finished (right now it's  
still WIP) GCCisms could be isolated / removed (because we would have  
a solid test base to test for possible regressions during this process).

Your help is always highly appreciated.

WBR,
Aleksey.