Re: [PATCH] mm: Don't warn if memdup_user fails

[Andrew Morton <akpm@linux-foundation.org>]

On Thu, 12 Jan 2012 13:19:54 -0800 (PST)
David Rientjes <rientjes@google.com> wrote:

> On Thu, 12 Jan 2012, Pekka Enberg wrote:
> 
> > I think you missed Andrew's point. We absolutely want to issue a
> > kernel warning here because ecryptfs is misusing the memdup_user()
> > API. We must not let userspace processes allocate large amounts of
> > memory arbitrarily.
> > 
> 
> I think it's good to fix ecryptfs like Tyler is doing and, at the same 
> time, ensure that the len passed to memdup_user() makes sense prior to 
> kmallocing memory with GFP_KERNEL.  Perhaps something like
> 
> 	if (WARN_ON(len > PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER))
> 		return ERR_PTR(-ENOMEM);
> 
> in which case __GFP_NOWARN is irrelevant.

If someone is passing huge size_t's into kmalloc() and getting failures
then that's probably a bug.  So perhaps we should add a warning to
kmalloc itself if the size_t is out of bounds, and !__GFP_NOWARN.

That might cause problems with those callers who like to call kmalloc()
in a probing loop with decreasing size_t.


But none of this will be very effective.  If someone is passing an
unchecked size_t into kmalloc then normal testing will not reveal the
problem because the testers won't pass stupid numbers into their
syscalls.