Re: [Qemu-devel] [PATCH 1/7] vfio: Remove unneeded union from VFIOContainer

[Thomas Huth <thuth@redhat.com>]

On 24/09/15 06:33, David Gibson wrote:
> Currently the VFIOContainer iommu_data field contains a union with
> different information for different host iommu types.  However:
>    * It only actually contains information for the x86-like "Type1" iommu
>    * Because we have a common listener the Type1 fields are actually used
> on all IOMMU types, including the SPAPR TCE type as well
>    * There's no tag in the VFIOContainer to tell you which union member is
> valid anyway.
> 
> In fact we now have a general structure for the listener which is unlikely
> to ever need per-iommu-type information, so this patch removes the union.
> 
> In a similar way we can unify the setup of the vfio memory listener in
> vfio_connect_container() that is currently split across a switch on iommu
> type, but is effectively the same in both cases.
> 
> The iommu_data.release pointer was only needed as a cleanup function
> which would handle potentially different data in the union.  With the
> union gone, it too can be removed.
> 
> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> ---
[...]
>          QLIST_FOREACH_SAFE(giommu, &container->giommu_list, giommu_next, tmp) {
> diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h
> index 9b9901f..fbbe6de 100644
> --- a/include/hw/vfio/vfio-common.h
> +++ b/include/hw/vfio/vfio-common.h
> @@ -59,22 +59,12 @@ typedef struct VFIOAddressSpace {
>  
>  struct VFIOGroup;
>  
> -typedef struct VFIOType1 {
> -    MemoryListener listener;
> -    int error;
> -    bool initialized;
> -} VFIOType1;
> -
>  typedef struct VFIOContainer {
>      VFIOAddressSpace *space;
>      int fd; /* /dev/vfio/vfio, empowered by the attached groups */
> -    struct {
> -        /* enable abstraction to support various iommu backends */
> -        union {
> -            VFIOType1 type1;
> -        };
> -        void (*release)(struct VFIOContainer *);
> -    } iommu_data;
> +    MemoryListener listener;
> +    int error;
> +    bool initialized;

Hmmm, maybe it's just bikeshed painting, but would it make sense to
rename the field with a "iommu_" prefix now, e.g. "iommu_error" instead
of "error", so that it is more clear that "error" refers to the IOMMU
stuff? (sorry for coming up with this now after suggesting to remove the
"iommu_data" container which made this clear ... but sometimes you have
to see the code first...)

 Thomas


>      QLIST_HEAD(, VFIOGuestIOMMU) giommu_list;
>      QLIST_HEAD(, VFIOGroup) group_list;
>      QLIST_ENTRY(VFIOContainer) next;
>