[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Xen-devel] [PATCH v7 10/10] xen/common: do not implicitly permit access to mapped I/O memory
- To: "Julien Grall" <julien.grall@xxxxxxxxxx>
- From: "Jan Beulich" <JBeulich@xxxxxxxx>
- Date: Mon, 26 May 2014 16:00:11 +0100
- Cc: Ian.Campbell@xxxxxxxxxxxxx, paolo.valente@xxxxxxxxxx, keir@xxxxxxx, stefano.stabellini@xxxxxxxxxxxxx, andrew.cooper3@xxxxxxxxxx, dario.faggioli@xxxxxxxxxx, Ian.Jackson@xxxxxxxxxxxxx, xen-devel@xxxxxxxxxxxxx, julien.grall@xxxxxxxxxx, etrudeau@xxxxxxxxxxxx, tim@xxxxxxx, Arianna Avanzini <avanzini.arianna@xxxxxxxxx>, viktor.kleinik@xxxxxxxxxxxxxxx
- Delivery-date: Mon, 26 May 2014 15:00:36 +0000
- List-id: Xen developer discussion <xen-devel.lists.xen.org>
>>> On 26.05.14 at 16:26, <julien.grall@xxxxxxxxxx> wrote:
> I'm still not convince, it's perfectly valid to:
> 1) call iomem_permission to set the permission
> 2) call memory_mapping
> 3) call iomem_permission to remove the permission
>
> In this case the memory mapping will be there without any granting.
> Hence, we won't be able to remove those mappings.
This would be a tool stack bug imo. And yes, this is possible, but no,
I don't think one should rely on this being possible (i.e. if we
introduced ways to find such mappings and refused the permission
removal, nothing should break).
Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel
|
