Since, with gc=yes, RELEASE is a noop, the --i never gets executed.
Thus, the loop is infinite. This seems to be the only place in the code
where there is side a effect of a expression in RELEASE. However, it
might be a good idea to have "#define RELEASE(object) ((id)object)"
instead of "#define RELEASE(object)" to protect against future problems
of this sort. The downside I can see to that is that there will be a
bunch more "statement with no effect" warnings.
*** Source-orig/NSArray.m Thu Mar 21 04:45:30 2002
--- Source/NSArray.m Fri May 10 04:14:54 2002
***************
*** 990,997 ****
objects[i] = [objects[i] copyWithZone: zone];
newArray = [[GSArrayClass allocWithZone: zone]
initWithObjects: objects count: count];
! while (i > 0)
! RELEASE(objects[--i]);
return newArray;
}
--- 990,997 ----
objects[i] = [objects[i] copyWithZone: zone];
newArray = [[GSArrayClass allocWithZone: zone]
initWithObjects: objects count: count];
! for (i = 0; i < count; i++)
! RELEASE(objects[i]);
return newArray;
}
_______________________________________________
Bug-gnustep mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-gnustep
