Re: [Xen-devel] [PATCH v3 7/7] xen/dom0: Add a dom0-iommu=none option

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 21.01.19 at 19:08, <andrew.cooper3@xxxxxxxxxx> wrote:
> On 17/01/2019 13:35, Jan Beulich wrote:
>>>>> On 16.01.19 at 10:00, <andrew.cooper3@xxxxxxxxxx> wrote:
>>> @@ -709,6 +709,12 @@ Controls for the dom0 IOMMU setup.
>>>      This option is enabled by default on x86 systems, and invalid on ARM
>>>      systems.
>>>  
>>> +*   The `none` option is intended for development purposes only, and skips
>>> +    certain safety checks pertaining to the correct IOMMU configuration for
>>> +    dom0 to boot.
>> Would you mind inserting "PVH" ahead of "dom0"?
> 
> That would result in an inaccurate description of the functionality. 
> check_hwdom_reqs() is not specific to PVH guests.

How is the paging_mode_translate() check your patch actually
amends not making this function effectively PVH-specific? Or
are you meaning to imply that some hypothetical future addition
to the checks affecting PV Dom0 is to be covered here as well?

>>> @@ -156,7 +159,7 @@ int iommu_domain_init(struct domain *d)
>>>  
>>>  static void __hwdom_init check_hwdom_reqs(struct domain *d)
>>>  {
>>> -    if ( !paging_mode_translate(d) )
>>> +    if ( iommu_hwdom_none || !paging_mode_translate(d) )
>>>          return;
>> Seeing the __hwdom_init, wouldn't it be better to restrict this
>> relaxation to Xen boot time created Dom0?
> 
> No, I don't think so.
> 
> That would complicate the change (which is already only for use by
> developers), and unnecessarily prohibit testing of the late hwdom paths.

Well, I don't fully agree, but okay then. Therefore if the answer
to the second question above is "yes", then
Acked-by: Jan Beulich <jbeulich@xxxxxxxx>

Jan



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel