> > > Would you be against a "mm->addr_limit"? That's what an approach > > > by Paulus implemented, and I was in the camp supporting that kind > > > of direction. > > > > On the other hand, this value is constant on some archs, and in those cases, > > wouldn't it be better for it to be a compile-time constant? > > How can it be constant? It would need to change when a set_fs(KERNEL_DS) > is performed.
It can be a constant on architectures that use a separate address space for the kernel, e.g. s390. In fact I don't set addr_limit at all, access_ok and friends always return 1. blue skies, Martin Martin Schwidefsky Linux for zSeries Development & Services IBM Deutschland Entwicklung GmbH
