Re: [PATCH v4 00/19] Intel IPU6 and IPU6 input system drivers

[Hans de Goede <hdegoede@redhat.com>]

Hi Sakari,

On 4/17/24 10:34 AM, Sakari Ailus wrote:
> Hi Hans,
> 
> On Wed, Apr 17, 2024 at 09:56:40AM +0200, Hans de Goede wrote:
>> Hi Sakari,
>> On 4/16/24 10:10 PM, Sakari Ailus wrote:
>>> Hello everyone,
>>>
>>> This patch series adds a driver for Intel IPU6 input system.
>>> IPU6 is the sixth generation of Imaging Processing Unit, it is a PCI
>>> device which can be found in some Intel Client Platforms. User can use
>>> IPU6 to capture images from MIPI camera sensors.
>>>
>>> IPU6 has its own firmware which exposes ABIs to driver, and communicates
>>> with CSE to do firmware authentication. IPU6 has its MMU hardware, so
>>> the driver sets up a page table to allow IPU6 DMA to access the system
>>> memory.
>>>
>>> IPU6 input system driver uses MC and V4L2 sub-device APIs besides V4L2.
>>>
>>> I can now capture images from ov01a10 and ov2740 sensors (including
>>> metadata from the latter).
>>>
>>> The series applies on top of the metadata patchset
>>> <URL:https://lore.kernel.org/linux-media/20240416193319.778192-1-sakari.ailus@linux.intel.com/T/#t>.
>>
>> Thank you for the new version!
>>
>> I assume that the posting of this new version means that you have solved
>> the stability issues where the laptop would freeze after sttreaming from
>> an ov2740 sensor with metadata once ?
>>
>> What about the unrelated ov2740 driver issue where the sensor would not
>> always start streaming for which you temporarily disabled runtime pm
>> for the sensor as a workaround any progress on that ?
> 
> I'm afraid these issues remain.

You mean both issues remain? I'm not that worried about the runtime-pm ov2740
issue, but if the lockup after streaming issue also remains that is a lot
more worrying.

I've been running an older version of this series without the metadata
support and that is pretty rock solid, so this seems to be caused by
enabling metadata support.

AFAIK the current out of tree solution with partly closed-source
userspace stack does not use metadata right /

Do you know if the Windows stack uses metadata capture from the sensor?

If neither the existing out of tree Linux stack nor the Windows stack
is using metadata capture then chances are we are actually hitting
hw/firmware bugs here. This would not be the first time that the Linux
community tries to enable a hw-feature not used by the factory installed
OS for the hw and ends up failing miserably because the feature was
never fully tested and turns out to be full of bugs.

IMHO if we cannot get the stability issue fixed real soon it would
be best to move forward with this patch series without adding
the metadata support. So basically drop patch 17/19 .

>> Do you have a git branch available with the metadata + this series
>> somewhere for easy testing ?  I would like to give this a test run on
>> my own IPU6 + ov2740 laptop.
> 
> Both of the sets can be found here:
> 
> <URL:https://git.linuxtv.org/sailus/media_tree.git/log/?h=ipu6>

Thank you.

Regards,

Hans