[RFC/PATCH] mm/futex: Fix futex writes on archs with SW tracking of dirty & young

[David Howells]

Peter Zijlstra <peterz at infradead.org> wrote:

> > What should nommu do anyways ? it's not like there's much it can do
> > right ? It should never even hit the fault path to start with ...
> 
> Something like the below makes a nommu arm config build.. David, is this
> indeed the correct thing to do for nommu?
> 
> ---
> Index: linux-2.6/mm/nommu.c
> ===================================================================
> --- linux-2.6.orig/mm/nommu.c
> +++ linux-2.6/mm/nommu.c
> @@ -190,6 +190,12 @@ int get_user_pages(struct task_struct *t
>  }
>  EXPORT_SYMBOL(get_user_pages);
>  
> +int fixup_user_fault(struct task_struct *tsk, struct mm_struct *mm,
> +		     unsigned long address, unsigned int fault_flags)
> +{
> +	BUG(); /* nommu should never call this */
> +}
> +
>  /**
>   * follow_pfn - look up PFN at a user virtual address
>   * @vma: memory mapping

Or perhaps send SEGV?  Can 'address' be bad at this point?

Can you inline this for the NOMMU case please?

David