Re: [PATCH V2 1/25] DOMCTL: Introduce new DOMCTL commands for vIOMMU support

[Lan Tianyu <tianyu.lan@intel.com>]

Hi Wei:
	Thanks for your review.

On 2017年08月17日 19:18, Wei Liu wrote:
> On Wed, Aug 09, 2017 at 04:34:02PM -0400, Lan Tianyu wrote:
>> diff --git a/xen/include/public/domctl.h b/xen/include/public/domctl.h
>> index ff39762..4b10f26 100644
>> --- a/xen/include/public/domctl.h
>> +++ b/xen/include/public/domctl.h
>> @@ -1149,6 +1149,56 @@ struct xen_domctl_psr_cat_op {
>>  typedef struct xen_domctl_psr_cat_op xen_domctl_psr_cat_op_t;
>>  DEFINE_XEN_GUEST_HANDLE(xen_domctl_psr_cat_op_t);
>>  
>> +/*  vIOMMU helper
>> + *
>> + *  vIOMMU interface can be used to create/destroy vIOMMU and
>> + *  query vIOMMU capabilities.
>> + */
>> +
>> +/* vIOMMU type - specify vendor vIOMMU device model */
>> +#define VIOMMU_TYPE_INTEL_VTD     (1u << 0)
> 
> Why use a bit when the types are mutually exclusive? Using a number
> should be fine?

Yes, will update.

> 
>> +
>> +/* vIOMMU capabilities */
>> +#define VIOMMU_CAP_IRQ_REMAPPING  (1u << 0)
>> +
>> +struct xen_domctl_viommu_op {
>> +    uint32_t cmd;
>> +#define XEN_DOMCTL_create_viommu          0
>> +#define XEN_DOMCTL_destroy_viommu         1
>> +#define XEN_DOMCTL_query_viommu_caps      2
>> +    union {
>> +        struct {
>> +            /* IN - vIOMMU type */
>> +            uint64_t viommu_type;
>> +            /* 
>> +             * IN - MMIO base address of vIOMMU. vIOMMU device models
>> +             * are in charge of to check base_address and length.
>> +             */
>> +            uint64_t base_address;
>> +            /* IN - Length of MMIO region */
>> +            uint64_t length;
>> +            /* IN - Capabilities with which we want to create */
>> +            uint64_t capabilities;
>> +            /* OUT - vIOMMU identity */
>> +            uint32_t viommu_id;
>> +        } create_viommu;
> 
> create should be fine.
> 

OK.

>> +
>> +        struct {
>> +            /* IN - vIOMMU identity */
>> +            uint32_t viommu_id;
>> +        } destroy_viommu;
> 
> destroy should be fine.
> 

OK.

-- 
Best regards
Tianyu Lan

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