Hi Laurent
On 15/12/2021 09:04, Laurent Pinchart wrote:
A month to the day! Sorry about the delay - more on that below...
> Hi Dan,
>
> On Tue, Dec 14, 2021 at 11:41:27PM +0000, Daniel Scally wrote:
>> On 14/12/2021 23:01, Laurent Pinchart wrote:
>>> On Tue, Dec 14, 2021 at 10:36:01PM +0000, Daniel Scally wrote:
>>>> On 14/12/2021 22:22, Laurent Pinchart wrote:
>>>>> On Mon, Dec 13, 2021 at 11:28:49PM +0000, Daniel Scally wrote:
>>>>>> Upon an async fwnode match, there's some typical behaviour that the
>>>>>> notifier and matching subdev will want to do. For example, a notifier
>>>>>> representing a sensor matching to an async subdev representing its
>>>>>> VCM will want to create an ancillary link to expose that relationship
>>>>>> to userspace.
>>>>>>
>>>>>> To avoid lots of code in individual drivers, try to build these links
>>>>>> within v4l2 core.
>>>>>>
>>>>>> Signed-off-by: Daniel Scally <djrscally@xxxxxxxxx>
>>>>>> ---
>>>>>> Changes since the rfc:
>>>>>>
>>>>>> - None
>>>>>>
>>>>>> drivers/media/v4l2-core/v4l2-async.c | 51 ++++++++++++++++++++++++++++
>>>>>> 1 file changed, 51 insertions(+)
>>>>>>
>>>>>> diff --git a/drivers/media/v4l2-core/v4l2-async.c b/drivers/media/v4l2-core/v4l2-async.c
>>>>>> index 0404267f1ae4..6575b1cbe95f 100644
>>>>>> --- a/drivers/media/v4l2-core/v4l2-async.c
>>>>>> +++ b/drivers/media/v4l2-core/v4l2-async.c
>>>>>> @@ -275,6 +275,45 @@ v4l2_async_nf_try_complete(struct v4l2_async_notifier *notifier)
>>>>>> static int
>>>>>> v4l2_async_nf_try_all_subdevs(struct v4l2_async_notifier *notifier);
>>>>>>
>>>>>> +static int
>>>>>> +__v4l2_async_create_ancillary_link(struct v4l2_async_notifier *notifier,
>>>>>> + struct v4l2_subdev *sd)
>>>>>> +{
>>>>>> + struct media_link *link;
>>>>>> +
>>>>>> + if (sd->entity.function != MEDIA_ENT_F_LENS &&
>>>>>> + sd->entity.function != MEDIA_ENT_F_FLASH)
>>>>>> + return -EINVAL;
>>>>>> +
>>>>>> + link = media_create_ancillary_link(¬ifier->sd->entity, &sd->entity,
>>>>> Is there a guarantee at this point that notifier->sd->entity has already
>>>>> been registered with the media_device ? That's done by
>>>>> media_device_register_entity() called from
>>>>> v4l2_device_register_subdev().
>>>> v4l2_async_match_notify() calls v4l2_device_register_subdev() before the
>>>> point that I've added the call to v4l2_async_try_create_links(), so I
>>>> think that's covered there.
>>> It calls it on sd, not notifier->sd. It's the latter that concerns me.
>> Ah, you're right of course...I guess in that case the notifier->sd would
>> get registered during the v4l2_async_match_notify() where the sensor
>> driver's subdev is sd, but I don't think there's any guarantee that that
>> would happen first...I haven't traced it through but my guess is that it
>> would depend on the order in which the ipu3-cio2, sensor and lens
>> controller drivers probed. I'll check to try and be sure how it works
>> tomorrow
> I was looking at media_device_register_entity(), and it sets
>
> INIT_LIST_HEAD(&entity->links);
> entity->num_links = 0;
> entity->num_backlinks = 0;
>
> If we create links before that, things may go bad.
It looks like there _is_ a guarantee of ordering actually. When the
ov8865's notifier is registered in v4l2_async_register_subdev_sensor(),
v4l2_async_nf_try_all_subdevs() is called against it, but at that point
v4l2_async_nf_find_v4l2_dev() won't find anything for the ov8865's
notifier even if the dw9719 has already probed and has it's async_subdev
waiting because the notifier has no parent and no directly assigned
v4l2_dev, so the function exits before trying to match anything (this
same logic guards all calls to v4l2_async_find_match()). Very shortly
after that v4l2_async_register_subdev() is called for the ov8865's
subdev which will match to ipu3-cio2's notifier. In
v4l2_async_match_notify() for that match the ipu3-cio2's notifier is
assigned as parent to the ov8865's notifier, but _after_
v4l2_device_register_subdev() is called for the ov8865. From that point
on v4l2_async_nf_find_v4l2_dev() will return a pointer and the matching
for the dw9719 will work correctly. So unless I've missed something, I
think it's ok.
This took me a long time to figure out, because I reset libcamera to
master for some reason and then totally forgot that I had done
that...which meant the auto-focus wasn't working when I tested it and I
convinced myself that my deliberate screwing of the probe ordering _did_
break it. After tearing my hair out for an embarrassing amount of time I
eventually figured out what I had done and got to the bottom of it -
sorry for the delay!
>
>>>>>> + MEDIA_LNK_FL_ENABLED |
>>>>>> + MEDIA_LNK_FL_IMMUTABLE);
>>>>>> +
>>>>>> + return IS_ERR(link) ? PTR_ERR(link) : 0;
>>>>>> +}
>>>>>> +
>>>>>> +/*
>>>>>> + * Setup links on behalf of the notifier and subdev, where it's obvious what
>>>>> s/Setup/Create/ ("link setup" refers to another operation, enabling and
>>>>> disabling links at runtime)
>>>> Yes, good point; although that too is a piece of terminology I find a
>>>> bit jarring to be honest; I would have named that function
>>>> media_entity_configure_link()
>>>>
>>>>>> + * should be done. At the moment, we only support cases where the notifier
>>>>>> + * is a sensor and the subdev is a lens.
>>>>> s/sensor/camera sensor/
>>>>> s/lens/lens controller/
>>>>>
>>>>>> + *
>>>>>> + * TODO: Setup pad links if the notifier's function is MEDIA_ENT_F_VID_IF_BRIDGE
>>>>>> + * and the subdev's is MEDIA_ENT_F_CAM_SENSOR
>>>>>> + */
>>>>>> +static int v4l2_async_try_create_links(struct v4l2_async_notifier *notifier,
>>>>>> + struct v4l2_subdev *sd)
>>>>>> +{
>>>>>> + if (!notifier->sd)
>>>>>> + return 0;
>>>>>> +
>>>>>> + switch (notifier->sd->entity.function) {
>>>>>> + case MEDIA_ENT_F_CAM_SENSOR:
>>>>>> + return __v4l2_async_create_ancillary_link(notifier, sd);
>>>>>> + default:
>>>>>> + return 0;
>>>>>> + }
>>>>>> +}
>>>>>> +
>>>>>> static int v4l2_async_match_notify(struct v4l2_async_notifier *notifier,
>>>>>> struct v4l2_device *v4l2_dev,
>>>>>> struct v4l2_subdev *sd,
>>>>>> @@ -293,6 +332,18 @@ static int v4l2_async_match_notify(struct v4l2_async_notifier *notifier,
>>>>>> return ret;
>>>>>> }
>>>>>>
>>>>>> + /*
>>>>>> + * Depending of the function of the entities involved, we may want to
>>>>>> + * create links between them (for example between a sensor and its lens
>>>>>> + * or between a sensor's source pad and the connected device's sink
>>>>>> + * pad)
>>>>> s/)/)./
>>>>>
>>>>>> + */
>>>>>> + ret = v4l2_async_try_create_links(notifier, sd);
>>>>>> + if (ret) {
>>>>>> + v4l2_device_unregister_subdev(sd);
>>>>>> + return ret;
>>>>>> + }
>>>>>> +
>>>>>> /* Remove from the waiting list */
>>>>>> list_del(&asd->list);
>>>>>> sd->asd = asd;
