Re: [x86] copy_from{to}_user question

From: H. Peter Anvin
Date: Fri Aug 21 2015 - 17:06:40 EST


On 08/20/2015 09:35 PM, Borislav Petkov wrote:
> On Thu, Aug 20, 2015 at 11:22:43AM -0700, H. Peter Anvin wrote:
>> There is a valid reason to do this, which is that currently
>> copy_{to,from}_user() effectively bypass SMAP as they don't verify that
>> the kernel pointer is actually a kernel pointer.
>
> Well, we do STAC before we copy but SMAP is checking for supervisor
> access to *user* data. But you say "kernel pointers" which is supervisor
> data. What am I missing?
>

What I'm saying is that we do do STAC, which *disables* SMAP. We have
to do that because one pointer is known to be a user space pointer.

However, we currently don't verify that the *other* pointer is kernel
space, which it is supposed to be (if not, we should be using
copy_in_user). We have to do this manually since we have to STAC which
means SMAP doesn't do anything at all. I believe it would be a good
idea to add such checks (and they would even benefit non-SMAP hardware.)

-hpa

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/