Re: [PATCH v3 2/2] tools: set event channel HVM parameters in libxenguest

[Andrew Cooper <amc96@xxxxxxxx>]

On 08/12/2021 15:57, Juergen Gross wrote:
> On 08.12.21 16:54, Andrew Cooper wrote:
>> On 08/12/2021 14:22, Juergen Gross wrote:
>>> On 08.12.21 14:43, Andrew Cooper wrote:
>>>> On 08/12/2021 08:47, Juergen Gross wrote:
>>>>> The HVM parameters for pre-allocated event channels should be set in
>>>>> libxenguest, like it is done for PV guests and for the pre-allocated
>>>>> ring pages.
>>>>>
>>>>> Suggested-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
>>>>> Signed-off-by: Juergen Gross <jgross@xxxxxxxx>
>>>>
>>>> I'm not sure that we have the concept of pre-allocated ring pages.
>>>>
>>>> For PV, we have:
>>>>
>>>>       dom->console_pfn = xc_dom_alloc_page(dom, "console");
>>>>       if ( dom->console_pfn == INVALID_PFN )
>>>>           return -1;
>>>>       xc_clear_domain_page(dom->xch, dom->guest_domid,
>>>>                            xc_dom_p2m(dom, dom->console_pfn));
>>>>
>>>> and for HVM, we have:
>>>>
>>>>       dom->console_pfn = special_pfn(SPECIALPAGE_CONSOLE);
>>>>       xc_clear_domain_page(dom->xch, dom->guest_domid,
>>>> dom->console_pfn);
>>>
>>> Isn't that a pre-allocation? The PFNs are fixed at boot time of the
>>> guest.
>>
>> Yeah, but "allocated in the library call we're making" is not the same
>> as "caller has to allocate and pass details in".
>>
>> I would not class the frames as "pre-allocated" in this context.
>> "allocated" sure, so perhaps "just like it is done for PV guests, and
>> the ring pages that libxenguest allocates" ?
>
> Fine with me.
>
> Should I send another round, or can this be changed when committing?

Fixed on commit.  No need to resend just for this.

Question is whether Anthony has any view, or whether my R-by is good enough?

~Andrew