[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Odd problem
From: |
David Chisnall |
Subject: |
Re: Odd problem |
Date: |
Thu, 28 Jul 2011 10:11:14 +0100 |
On 28 Jul 2011, at 08:37, Fred Kiefer wrote:
> Now you are on your own. I would suggest that you write out the delegate when
> ever it gets set, that way you may compare the address with the one reported
> by gdb when the segmentation fault happens. Then you will know which object
> gets removed without clearing the delegate. Anyway, this problem is so
> complicated as it spans over different frameworks. Good luck with tracking it
> down.
I'd also recommend:
- Run the static analyser on your code. This will pick up unbalanced releases.
- Enabling zombines - this will print a message telling you when you send a
message to a dealloc'd object.
- Once you know what its class is, add a breakpoint in its dealloc method and
see where it is being released. Also add a custom allocWithZone: (that just
calls super) and a breakpoint on it so that you can see where it is created.
David
-- Sent from my STANTEC-ZEBRA
- Re: Odd problem, (continued)
- Re: Odd problem, Riccardo Mottola, 2011/07/25
- Re: Odd problem, Fred Kiefer, 2011/07/25
- Re: Odd problem, Riccardo Mottola, 2011/07/25
- Re: Odd problem, Germán Arias, 2011/07/25
- Re: Odd problem, David Chisnall, 2011/07/25
- Re: Odd problem, Fred Kiefer, 2011/07/25
- Re: Odd problem, Germán Arias, 2011/07/26
- Re: Odd problem, Fred Kiefer, 2011/07/27
- Re: Odd problem, Germán Arias, 2011/07/28
- Re: Odd problem, Fred Kiefer, 2011/07/28
- Re: Odd problem,
David Chisnall <=