User forums > Nightly builds
The 18 June 2006 build is out.
killerbot:
--- Quote from: royalbox on June 20, 2006, 08:08:59 pm ---Right, I tried it with gdb. As before, I clicked on the 'open' icon and got:
--- Quote ---Program received signal SIGSEGV, Segmentation fault.
0x77ea3c00 in RpcRaiseException () from C:\WINDOWS\system32\rpcrt4.dll
--- End quote ---
I then typed
--- Code: ---bt all
--- End code ---
and got:
--- Quote ---No symbol "all" in current context.
--- End quote ---
--- End quote ---
just use the debug version (don't let gdb run it) and let it crash, see if then an rpt is created ?
royalbox:
@Michael
Thanks for all that info, I'll keep a link to it in case I change my mind one of these days and want to compile CB.
@killerbot
No gdb, okay.
This is what I did then. I extracted the debug version to a folder on my desktop. I added the required dll's -- mingwm10.dll and the other one. I deleted the codeblocks folder from documents and settings\.... I ran codeblocks directly from the exe. After the splash screen, compiler selection dialog, tips dialog and file association dialog, code::blocks opened. I clicked on the open icon and it brought up the expected "whoa!" "access violation" dialog. I clicked on abort, the dialog closed and the 'open' dialog opened as if nothing had happened. I closed the 'open' dialog, closed code::blocks and searched for any kind of crash report. All that I could find was the "cb-crash-recover" folder that was created.
That's it. Unless you have any other suggestions to try.
Perhaps some of the other posters who had the same problem will try your debug build and see if they have any luck, I hope so.
thomas:
I have changed the crash handler to explicitely only handle two exceptions now:
segmentation fault, and illegal instruction.
Although I cannot reproduce the problem and don't know what is causing it for you, this might solve the problem.
On the other hand, this means Code::Blocks may crash again on certain events (because they are not caught any more), but a segmentation fault due to an uninitialised pointer is the most frequently occurring bug anyway (well over 95% of all crashes). So I guess handling this and illegal instructions is good enough.
killerbot:
see my post in the nightly build of the 20th, I also had crashes and after abort i could continue ...
royalbox:
Ok, thanks everyone for all your help and hard work. I'm only sorry my testing couldn't give you the answer.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version