qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [RFC 1/3] memory: add MemTxAttrs to translate function


From: David Gibson
Subject: Re: [Qemu-devel] [RFC 1/3] memory: add MemTxAttrs to translate function
Date: Wed, 6 Jun 2018 10:11:18 +1000
User-agent: Mutt/1.9.5 (2018-04-13)

On Tue, Jun 05, 2018 at 02:38:31PM +0100, Peter Maydell wrote:
> On 5 June 2018 at 14:19, Peter Xu <address@hidden> wrote:
> > Add a new MemTxAttrs parameter to the IOMMUMemoryRegionClass.translate()
> > function, which takes some extra context of the translation request.
> >
> > Signed-off-by: Peter Xu <address@hidden>
> > ---
> >  include/exec/memory.h    | 5 ++++-
> >  exec.c                   | 2 +-
> >  hw/alpha/typhoon.c       | 3 ++-
> >  hw/arm/smmuv3.c          | 2 +-
> >  hw/dma/rc4030.c          | 6 ++++--
> >  hw/i386/amd_iommu.c      | 2 +-
> >  hw/i386/intel_iommu.c    | 6 ++++--
> >  hw/ppc/spapr_iommu.c     | 3 ++-
> >  hw/s390x/s390-pci-bus.c  | 6 ++++--
> >  hw/sparc/sun4m_iommu.c   | 3 ++-
> >  hw/sparc64/sun4u_iommu.c | 3 ++-
> >  memory.c                 | 3 ++-
> >  12 files changed, 29 insertions(+), 15 deletions(-)
> >
> > diff --git a/include/exec/memory.h b/include/exec/memory.h
> > index eb2ba06519..6b0ced554d 100644
> > --- a/include/exec/memory.h
> > +++ b/include/exec/memory.h
> > @@ -234,9 +234,12 @@ typedef struct IOMMUMemoryRegionClass {
> >       * @iommu: the IOMMUMemoryRegion
> >       * @hwaddr: address to be translated within the memory region
> >       * @flag: requested access permissions
> > +     * @attrs: MemTxAttrs that was bound to the translation
> > +     *         operation. If flag==IOMMU_NONE, this field is
> > +     *         meaningless
> >       */
> 
> This won't work, because the "allow TCG transactions to pass
> through an IOMMU" code needs to pass IOMMU_NONE. IOMMU_NONE
> means "give me all the permissions info, and don't shortcut it".
> For TCG we get the info, and then cache it and use it for
> subsequent transactions, both read and write, regardless of
> whether the first one was a read or a write.

Likewise the "replay" operations needed for hotplug use IOMMU_NONE.

It also seems like it's kind of abusing the meaning of "AddressSpace",
if other factors are influencing how the translation is done.

-- 
David Gibson                    | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au  | minimalist, thank you.  NOT _the_ _other_
                                | _way_ _around_!
http://www.ozlabs.org/~dgibson

Attachment: signature.asc
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]