Re: OMAP4 support

[Gary Thomas <gary@chez-thomas.org>]

Ping?  Can anyone help me with these issues?

On 2012-07-15 14:24, Gary Thomas wrote:
> On 2012-07-15 08:31, Sergio Aguirre wrote:
>> Hi Gary,
>>
>> On Fri, Jul 13, 2012 at 5:24 AM, Gary Thomas <gary@mlbassoc.com> wrote:
>>> On 2012-07-12 20:30, Sergio Aguirre wrote:
>>>>
>>>> Hi Gary,
>>>>
>>>> On Tue, Jul 10, 2012 at 2:31 PM, Gary Thomas <gary@mlbassoc.com> wrote:
>>>>>
>>>>> On 2012-07-10 11:05, Chris Lalancette wrote:
>>>>>>
>>>>>>
>>>>>> On Tue, Jul 10, 2012 at 9:41 AM, Gary Thomas <gary@mlbassoc.com> wrote:
>>>>>>>
>>>>>>>
>>>>>>> I'm looking for video support on OMAP4 platforms.  I've found the
>>>>>>> PandaBoard camera project
>>>>>>> (http://www.omappedia.org/wiki/PandaBoard_Camera_Support)
>>>>>>> and this is starting to work.  That said, I'm having some
>>>>>>> issues with setting up the pipeline, etc.
>>>>>>>
>>>>>>> Can this list help out?
>>>>>>
>>>>>>
>>>>>>
>>>>>> I'm not sure exactly what kind of cameras you want to get working, but
>>>>>> if you are looking to get CSI2 cameras going through the ISS, Sergio
>>>>>> Aguirre has been working on support.  He also works on the media-ctl
>>>>>> tool, which is used for configuring the media framework pipeline.  The
>>>>>> latest versions that I am aware of are here:
>>>>>>
>>>>>> git://gitorious.org/omap4-v4l2-camera/omap4-v4l2-camera.git
>>>>>
>>>>>
>>>>>
>>>>> Yes, this is the tree I've been working with (pointed to by the page I
>>>>> mentioned).
>>>>>
>>>>> My kernel can see the camera OV5650 and set up the pipeline.  I am able
>>>>> to
>>>>> grab
>>>>> the raw SGRBG10 data but I'd like to get the ISS to convert this to a
>>>>> more
>>>>> usable
>>>>> UYVY format.  Here's what I tried:
>>>>>     media-ctl -r
>>>>>     media-ctl -l '"OMAP4 ISS CSI2a":1 -> "OMAP4 ISS ISP IPIPEIF":0 [1]'
>>>>>     media-ctl -l '"OMAP4 ISS ISP IPIPEIF":1 -> "OMAP4 ISS ISP IPIPEIF
>>>>> output":0 [1]'
>>>>>     media-ctl -f '"ov5650 3-0036":0 [SGRBG10 2592x1944]'
>>>>>     media-ctl -f '"OMAP4 ISS CSI2a":0 [SGRBG10 2592x1944]'
>>>>>     media-ctl -f '"OMAP4 ISS ISP IPIPEIF":0 [SGRBG10 2592x1944]','"OMAP4
>>>>> ISS
>>>>> ISP IPIPEIF":1 [UYVY 2592x1944]'
>>>>>
>>>>> Sadly, I can't get the IPIPEIF element to take SGRGB10 in and put UYVY
>>>>> out
>>>>> (my reading
>>>>> of the manual implies that this _should_ be possible).  I always see this
>>>>> pipeline setup:
>>>>> - entity 5: OMAP4 ISS ISP IPIPEIF (3 pads, 4 links)
>>>>>               type V4L2 subdev subtype Unknown
>>>>>               device node name /dev/v4l-subdev2
>>>>>           pad0: Input [SGRBG10 2592x1944]
>>>>>                   <- 'OMAP4 ISS CSI2a':pad1 [ACTIVE]
>>>>>                   <- 'OMAP4 ISS CSI2b':pad1 []
>>>>>           pad1: Output [SGRBG10 2592x1944]
>>>>>                   -> 'OMAP4 ISS ISP IPIPEIF output':pad0 [ACTIVE]
>>>>>           pad2: Output [SGRBG10 2592x1944]
>>>>>                   -> 'OMAP4 ISS ISP resizer':pad0 []
>>>>>
>>>>> Am I missing something?  How can I make this conversion in the ISS?
>>>>
>>>>
>>>> The core problem is that, i haven't published any support for
>>>> RAW10->YUV conversion,
>>>> which is part of the IPIPE module (not the IPIPEIF, like you mention). I
>>>> had
>>>> some patches, but sadly it is unfinished work. :/
>>>>
>>>> Now, there's a main non-technical problem... I no longer work at TI
>>>> since end of June
>>>> this year, and I don't have the right HW setup available anymore.
>>>> Those sensors were
>>>> company's asset, and I couldn't keep any.
>>>>
>>>> Now, we can make this work with cooperation of someone who has the right
>>>> setup,
>>>> and me sharing my patches and some advice on my experience.
>>>>
>>>> What do you think?
>>>>
>>>>>
>>>>> Note: if this is not the appropriate place to ask these questions, please
>>>>> redirect me (hopefully to a useful list :-)
>>>>
>>>>
>>>> As I'm the main person who has been actively developing this, I'm your
>>>> guy to ask questions :).
>>>>
>>>> By the way, this development has been my initiative the whole time,
>>>> and not an official
>>>> TI objective, so, to be honest, asking TI for official support won't
>>>> help much right now.
>>>
>>>
>>> Tell me how I can help make this happen.  I'll be glad to apply patches,
>>> figure out bugs, etc, I just need a little help with getting started.
>>> I have access to the hardware and it's really important that I make some
>>> progress on this soon.
>>>
>>> Can you share your RAW10->YUV patches and some guidance on how to proceed?
>>
>> Sure. I just pushed an internal branch I had, named: "devel-ISPSUPPORT-IPIPE",
>> please take that as a base.
>>
>> And please try these commands:
>>
>> media-ctl -r -l '"OMAP4 ISS CSI2a":1 -> "OMAP4 ISS ISP IPIPEIF":0
>> [1]','"OMAP4 ISS ISP IPIPEIF":2 -> "OMAP4 ISS ISP IPIPE":0
>> [1]','"OMAP4 ISS ISP IPIPE":1 -> "OMAP4 ISS ISP resizer":0
>> [1]','"OMAP4 ISS ISP resizer":1 -> "OMAP4 ISS ISP resizer a output":0
>> [1]'
>>
>> media-ctl -f '"ov5650 3-0036":0 [SGRBG10 2592x1944]','"OMAP4 ISS
>> CSI2a":0 [SGRBG10 2592x1944]','"OMAP4 ISS ISP IPIPEIF":0 [SGRBG10
>> 2592x1944]','"OMAP4 ISS ISP IPIPE":0 [SGRBG10 2592x1944]','"OMAP4 ISS
>> ISP resizer":0 [UYVY 2592x1944]'
>>
>> yavta /dev/video3 -c4 -n1 -s2592x1944 -fUYVY -Fov5650_2592x1944_UYVY_8bpp.yuv
>
> With the new branch, I am able to set up the pipeline for UYVY.  That part's good.
>
> However, just like before, with either RAW10 or UYVY, the grab process does not
> start more times than it does (it only starts about 1 out of 10 tries).  If I
> just ^C and try again, it may start, it may not.
>
> The single time I was able to get the UYVY capture to work, I got an error after
> the first frame:
>
> # grab-uyvy
> Device /dev/video3 opened.
> Device `OMAP4 ISS ISP resizer a output' on `media' is a video capture device.
> Video format set: UYVY (59565955) 2592x1944 buffer size 10077696
> Video format: UYVY (59565955) 2592x1944 buffer size 10077696
> [  622.039733] omap4iss omap4iss: dma_alloc_coherent of size 10077696 failed
> 1 buffers requested.
> length: 10077696 offset: 0
> Buffer 0 mapped at address 0xb64a2000.
> 0 (0) [-] 0 1007[  623.131347] omap4iss omap4iss: RSZ Err: FIFO_IN_BLK:0, FIFO_OVF:1,
> 7696 bytes 623.118776 1342383344.713594 0.001 fps
> [  623.297790] omap4iss omap4iss: RSZ Err: FIFO_IN_BLK:1, FIFO_OVF:0,
>
> What can I look at to figure out why the capture is hanging?  I already enabled
> a debug print in the ISR and see that only fires when it works.  There are no
> interrupts from the CSI2a when it fails.
>
> Any hints as to where & what to check (print registers, etc) would be a big help.
>
> Thanks
>
>>
>>>
>>> I have been able to capture RAW10 data, but often the whole thing just sits
>>> there (hangs).  Restarting the process sometimes works, sometimes not.
>>> Looking
>>> at the registers and the actual signals on a scope do not show any
>>> difference
>>> that we can find.  Any ideas what might cause this?  Have you seen it as
>>> well?
>>
>> Can you please try again with the before mentioned branch? The branch you
>> were using didn't have some changes, so maybe this new one would take
>> care of that.
>>
>>>
>>> Thanks for the help - Please let me know how I can get this working...
>>
>> Well, thanks for the patience!
>>
>> Regards,
>> Sergio
>

-- 
Gary Thomas