Re: [PATCH v3 3/3] xen/pci: Refactor MSI code that implements MSI functionality within XEN

[Jan Beulich <jbeulich@xxxxxxxx>]

On 28.04.2021 13:06, Roger Pau Monné wrote:
> On Mon, Apr 26, 2021 at 05:21:27PM +0100, Rahul Singh wrote:
>> MSI code that implements MSI functionality to support MSI within XEN is
>> not usable on ARM. Move the code under CONFIG_PCI_MSI_INTERCEPT flag to
>> gate the code for ARM.
>>
>> Currently, we have no idea how MSI functionality will be supported for
>> other architecture therefore we have decided to move the code under
>> CONFIG_PCI_MSI_INTERCEPT. We know this is not the right flag to gate the
>> code but to avoid an extra flag we decided to use this.
>>
>> No functional change intended.
>>
>> Signed-off-by: Rahul Singh <rahul.singh@xxxxxxx>
> 
> I think this is fine, as we don't really want to add another Kconfig
> option (ie: CONFIG_PCI_MSI) for just the non explicitly intercept MSI
> code.

While a separate config option may be excessive, keying it to the
wrong one is not desirable imo. If we want to avoid having PCI_MSI,
then keeping respective bits guarded by X86 would look better to me.
I'm not convinced though that doing the separation properly right
away (see also the XSM changes in patch 2 which imo belong here)
isn't going to be better in the long run, and hence introducing
HAS_PCI_MSI right away isn't the way to go.

Jan