David Matlack <dmatlack@xxxxxxxxxx> writes:
> On Mon, Aug 18, 2025 at 4:41 PM Vinicius Costa Gomes
> <vinicius.gomes@xxxxxxxxx> wrote:
>> David Matlack <dmatlack@xxxxxxxxxx> writes:
>> > +
>> > +static int dsa_probe(struct vfio_pci_device *device)
>> > +{
>> > + if (!vfio_pci_device_match(device, PCI_VENDOR_ID_INTEL,
>> > + PCI_DEVICE_ID_INTEL_DSA_SPR0))
>>
>> What are you thinking about adding support for multiple device ids?
>
> I haven't given it much thought yet. But we could definitely support
> fancier device matching (e.g. multiple acceptable device ids) if/when
> a use-case for that arises.
>
Fair enough. I just wanted to bring this up (hoping that's more a matter
of "when" than "if" :-))
>> > +static int dsa_completion_wait(struct vfio_pci_device *device,
>> > + struct dsa_completion_record *completion)
>> > +{
>> > + u8 status;
>> > +
>> > + for (;;) {
>> > + dsa_check_sw_err(device);
>> > +
>> > + status = READ_ONCE(completion->status);
>> > + if (status)
>> > + break;
>> > +
>> > + usleep(1000);
>>
>> Another minor/thing to think about: using umonitor/umwait.
>
> Thanks for the tip, I hadn't considered that. But I think for this
> driver, keeping things as simple as possible is best. This code is
> only used for testing so I don't think we care enough about efficiency
> to justify using unmonitor/umwait here.
Yeah, agreed.
Cheers,
--
Vinicius
