On Mon, 19 Jul 2004 cgd@xxxxxxxxxxxx wrote: > At Mon, 19 Jul 2004 14:22:29 +0000 (UTC), "Maciej W. Rozycki" wrote: > > (this has led to an address shift enlarging the > > patch significantly, unfortunately). > > maybe add nops to pad, instead? Well, addresses have been shifted down, so you'd have to remove something. I suppose the break tests could be moved to the end of file, but I fail to see an advantage. Test cases that embed addresses in output patterns are doomed to suffer from such changes. Maciej
