On 2017年08月23日 15:45, Roger Pau Monné wrote:
> On Wed, Aug 09, 2017 at 04:34:11PM -0400, Lan Tianyu wrote:
>> From: Chao Gao <chao....@intel.com>
>>
>> If guest is configured to have a vIOMMU, create it during domain
>> construction.
>>
>> Signed-off-by: Chao Gao <chao....@intel.com>
>> Signed-off-by: Lan Tianyu <tianyu....@intel.com>
>> ---
>> tools/libxl/libxl_x86.c | 28 ++++++++++++++++++++++++++++
>> 1 file changed, 28 insertions(+)
>>
>> diff --git a/tools/libxl/libxl_x86.c b/tools/libxl/libxl_x86.c
>> index 455f6f0..ace20e5 100644
>> --- a/tools/libxl/libxl_x86.c
>> +++ b/tools/libxl/libxl_x86.c
>> @@ -341,8 +341,36 @@ int libxl__arch_domain_create(libxl__gc *gc,
>> libxl_domain_config *d_config,
>> if (d_config->b_info.type == LIBXL_DOMAIN_TYPE_HVM) {
>
> I would rather change this check so it's:
>
> d_config->b_info.type != LIBXL_DOMAIN_TYPE_PV
>
> Is there any reason why PVH guests shouldn't get a vIOMMU?
No, but we current only support vIOMMU for HVM guest and don't know how
PVH guest enumerates vIOMMU without ACPI DMAR table.
--
Best regards
Tianyu Lan
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
