Does i.MX 8X support V4L2_FIELD_ALTERNATE with MIPI CSI-2?

[petero5]

Hi, we are using i.MX 8QuadXPlus with ADV7280A-M over MIPI CSI-2. And the i2c adv7180 driver.

Using VIDIOC_S_FMT with fmt.fmt.pix_mp.field = V4L2_FIELD_ALTERNATE

Following VIDIOC_DQBUF:

buf.field is 1 V4L2_FIELD_NONE...

and buf.sequence increments each time (contrary to "In V4L2_FIELD_ALTERNATE mode the top and bottom field have the same sequence number.") - https://www.kernel.org/doc/html/latest/userspace-api/media/v4l/buffer.html#struct-v4l2-buffer

We are getting 50fps, 720 x 288, so we are receiving interlaced frames.

But the meta data in the v4l2_buffer isn't as expected for V4L2_FIELD_ALTERNATE.

Is there something more we need to do to enable V4L2_FIELD_ALTERNATE support in the drivers?

Thank you
-- Peter

[joanxie]

I sent a patch to you for reference, this is just an example, not official released version, so I couldn't share it on the public community, pls check your mail

[ToarteFretter]

Hi @joanxie,

We're using the same ADV7280 chip and worked closely quite some time ago with NXP on this de-interlacing issue on the i.MX8QM, and basically came to the conclusion it's impossible to get the odd/even detection to work properly or to get automatically right in case the odd/even cadence is wrong.

So we're very interested in this patch!

Can you share this please?

Thank you very much.

Hi @petero5, were you able to properly de-interlace the odd/even fields from the ADV7280?

Best regards.

[joanxie]

pls submit a ticket, otherwise I couldn't mail to you direclty

https://www.nxp.com/support/support:SUPPORTHOME?tid=sbmenu

[ToarteFretter]

Hi @joanxie,

I just submitted a ticket, Case Number 00715976.

Thanks!

Best regards.

[petero5]

Similar unresolved issue:
https://community.nxp.com/t5/i-MX-Processors/Trouble-with-deinterlacing-NTCS-PAL-video-with-an-IMX8M...