On Wed, Jul 05, 2000 at 05:06:02PM -0400, Rick Scott wrote:
> But it's not exactly a resource offset once compiled, right?? After all, we are
> looking for a (int)Cardinal < 0, which means a Cardinal > 2147483647 (assuming
> 4 byte int's and Cardinal's).
Yup, the MAXINT issue is also interesting here ...
This might also look like a portability problem (-> alpha systems):
Cardinal is 'unsigned int' in X11, i.e. if it's going to
be used for size_t one had to face serious trouble.
But did omitting this check, i.e. take the else{} part fix test7 for you??
>I suspect we have some sort of _very_ serious corruption going on. Try this...
Interesting point.
Given the stability of LT in general we either rarely run across this
or just by chance it doesn't hurt often?! ;-)
>gdb ./test7
>break main
>run
>break BaseClass.c:_XmTransformSubResources
>cont
>where
>
>You end up at _XmGetClassExtensionPtr () at BaseClass.c:1903!!!!!!!
>
>What gives?????
Hmm, are you sure you have disabled optimization for compiling the lib?
Actually that's an issue I always have to fight with: I have two
'scripts' to run configure for my Production and Debug trees but
for some strange reasons I often have -O!=0 used in my Debug/ :-(
--
Alexander Mai
[EMAIL PROTECTED]
