Re: [Xen-devel] [PATCH] xen: workaround for 64-bit size/alignment bitops

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Mon, 2014-04-28 at 10:10 +0100, David Vrabel wrote:
> On 28/04/14 09:47, Pranavkumar Sawargaonkar wrote:
> > On 27 April 2014 14:39, Vladimir Murzin <murzin.v@xxxxxxxxx> wrote:
> >> Xen assumes that bit operations are able to operate on 32-bit size and
> >> alignment. For arm64 bitops are based on atomic exclusive load/store
> >> instructions to guarantee that changes are made atomically. However, these
> >> instructions require that address to be aligned to the data size. Because, 
> >> by
> >> default, bitops operates on 64-bit size it implies that address should be
> >> aligned appropriately. All these lead to breakage of Xen assumption for 
> >> bitops
> >> properties.
> >>
> >> As a workaround address is aligned forcefully along with adjustment for bit
> >> position.
> >>
> [...]
> > 
> > I have tried this patch on XGENE and works fine, I have managed to get
> > dom0 and domU working with it.
> > 
> > Tested-by: Pranavkumar Sawargaonkar <pranavkumar@xxxxxxxxxx>
> 
> Thanks.  Are there arm64 platforms that support Xen currently available
> in 3.15?  Or can this wait until 3.16?

The foundation model and fastmodels are supported by 3.15. Xgene is
partially available in 3.15 I think. This should be fixed in 3.15
please, and ideally be queued for backport to whichever version
introduced the FIFO evtchn stuff.

Ian.


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel