Re: [PATCH v9 15/28] ARM: vITS: provide access to struct pending_irq

[Julien Grall <julien.grall@arm.com>]

On 22/05/17 17:50, Andre Przywara wrote:
> Hi,

Hi Andre,

> On 17/05/17 16:35, Julien Grall wrote:
>>> +    }
>>> +    spin_unlock(&d->arch.vgic.its_devices_lock);
>>> +
>>> +    return pirq;
>>> +}
>>> +
>>> +struct pending_irq *gicv3_its_get_event_pending_irq(struct domain *d,
>>> +                                                    paddr_t
>>> vdoorbell_address,
>>> +                                                    uint32_t vdevid,
>>> +                                                    uint32_t veventid)
>>
>> s/veventid/eventid/
>>
>>> +{
>>> +    return get_event_pending_irq(d, vdoorbell_address, vdevid,
>>> veventid, NULL);
>>> +}
>>
>> This wrapper looks a bit pointless to me. Why don't you directly expose
>> get_event_pending_irq(...)?
>
> I don't want to expose host_lpi in the exported function, because it's
> of no need for the callers and rather cumbersome for them to pass NULL
> or the like. But then the algorithm to find host_lpi and pirq is
> basically the same, so I came up with this joint static function and an
> exported wrapper, which hides the host_lpi.
> And there is one user (in gicv3_assign_guest_event()) which needs both,
> so ...
> If you can think of a better way to address this, I am all ears.

It is not that bad to pass NULL everywhere. We already have some other 
functions like that.

How about adding the wrapper as a static inline in the header?

Cheers,

-- 
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel