Andrew Haley a écrit :
Are you using optimization?
No, I'm building with those CFLAGS : -g -ggdb -Wall -save-temps
-fno-strict-aliasing
I think can see at least one bug in this function.
((CARD32 *)data)[i] = ((long *)retval)[i];
You're absolutely right, CARD32 is defined as an "unsigned long" on x86
but as an "unsigned int" on x86_64. Changing this does not however
change anything in how the Xlib call done just before behaves.
Try compiling with -fno-strict-aliasing.
Doesn't change anything. No additional warnings nor any change in behavior.
--
Rémi Cardona
LRI, INRIA
remi.cardona@xxxxxx
remi@xxxxxxxxxx
