On Mon, 31 Jan 2005 10:23:54 +0100 Andi Kleen <[EMAIL PROTECTED]> wrote:
> > However, I like the mm->max_addr idea because that could be used > > for the mmap()/munmap()/mremap() sanity checks as well instead of > > bogus TASK_SIZE. > > Hmm, but in process context it is not bogus is it? I guess you're suggesting that there could be times when mm->max_addr and the current thread's address space disposition are out of sync? That is certainly something to check out. And I am rather certain that it is flush_old_exec() which should transition mm->max_addr. It is the only spot where the address space API logically changes. More precisely, this occurs when exec_mmap() completes successfully, so maybe it should happen there and in that case.
