Concern about AMD IOMMU HPET msi remapping

Concern about AMD IOMMU HPET msi remapping

[Andrew Cooper]

Hello,

In the committed patch
http://xenbits.xen.org/hg/staging/xen-unstable.hg/rev/9025a10c0561

The bdf parameter to update_intremap_entry_from_msi_msg() is declared as
u8.  The single caller of this function passes bdf as an int (which is
only ever filled with a u16), while the this function passes bdf to
get_{dma,intremap}_requestor_id(), which expect u16's

Is there not a truncation issue by doing this, or am I missing something?

-- 
Andrew Cooper - Dom0 Kernel Engineer, Citrix XenServer
T: +44 (0)1223 225 900, http://www.citrix.com

Re: Concern about AMD IOMMU HPET msi remapping

[Jan Beulich]

>>> On 19.11.12 at 14:33, Andrew Cooper <andrew.cooper3@citrix.com> wrote:
> Hello,
> 
> In the committed patch
> http://xenbits.xen.org/hg/staging/xen-unstable.hg/rev/9025a10c0561 
> 
> The bdf parameter to update_intremap_entry_from_msi_msg() is declared as
> u8.  The single caller of this function passes bdf as an int (which is
> only ever filled with a u16), while the this function passes bdf to
> get_{dma,intremap}_requestor_id(), which expect u16's
> 
> Is there not a truncation issue by doing this, or am I missing something?

Oh, indeed - thanks for spotting. Will fix this right away.

Jan