Re: Possible GCC Bug?
Re: Possible GCC Bug?
- Subject: Re: Possible GCC Bug?
- From: "B.J. Buchalter" <email@hidden>
- Date: Tue, 8 Apr 2008 16:19:55 -0400
On Apr 8, 2008, at 12:53 PM, Jack Repenning wrote:
On Apr 8, 2008, at 9:31 AM, Greg Knight wrote:
I've been away from 'C' for over ten years, and just come back to
port a legacy app to the MAC. The following code is misbehaving
under the XCode 3.0 Debug configuration (GCC4). I'd appreciate any
suggestions.
// int fred has value 13
DisposeWindow(WinData->window); // Carbon call
// int fred now has value 27
Possible, but vastly more likely the code has a heretofore
undiscovered memory management bug. Check the documentation for
"Guard Malloc" in Xcode (open documentation window, search for guard
malloc; probably start with the "Memory Usage Performance
Guidelines" document, page "Enabling the Malloc Debugging
Features"). Then use Xcode's "Enable Guard Malloc" menu pick (in
Run), and see what you can find.
You might also check whatever comes just before fred, and whatever
comes just after it, wherever it is fred is declared, to see if
they're getting trounced as well.
I don't think GaurdMalloc will help much here, but using the Xcode/GDB
"watch variable" command (to enable a watchpoint) will. Set a
watchpoint on "fred" and Xcode will stop in the debugger on the
precise machine instruction that changes "fred".
Very cool.
B.J. Buchalter
Metric Halo
http://www.mhlabs.com
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden