DSS2 questions

DSS2 questions

[Gary Thomas]

Tomi,

I've been following your DSS2 work, with reasonable success.
However, when I tried to pull from your repo today, I got
a ton of merge errors.  I cloned your tree a [little] while
back (it seems it was 2009-06-16, but where did the time go?)

Here's what I did:
  git clone http://www.bat.org/~tomba/git/linux-omap-dss.git
    HEAD=5bd374f9b199d46ae434489d56615b240de5a6d7
  git checkout -b my_branch origin/master
    ...
When I went to pull into my_branch, I got the merge errors.
Any clues what I did wrong?  How could/should I have managed
this better?

n.b. you earlier told me that git was better than sliced
bread; I'm still waiting to be shown the light :-)

Second question; I've set up my system much the same as the
boards you are working on, with an LCD and TV outputs.  I'm
a bit confused as to how I configure the overlays and managers
to be able to send YUV data directly through to either the
LCD or TV.  Can you explain this?

Thanks

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: DSS2 questions

[Tomi Valkeinen]

Hi,

ext Gary Thomas wrote:
> Tomi,
> 
> I've been following your DSS2 work, with reasonable success.
> However, when I tried to pull from your repo today, I got
> a ton of merge errors.  I cloned your tree a [little] while
> back (it seems it was 2009-06-16, but where did the time go?)
> 
> Here's what I did:
>   git clone http://www.bat.org/~tomba/git/linux-omap-dss.git
>     HEAD=5bd374f9b199d46ae434489d56615b240de5a6d7
>   git checkout -b my_branch origin/master
>     ...
> When I went to pull into my_branch, I got the merge errors.
> Any clues what I did wrong?  How could/should I have managed
> this better?

The problem is that I have rebased my tree so that I end up with a clean 
set of patches that can be posted to mailing lists. And rebasing means 
changing history, which means that you cannot merge it normally.

Perhaps I should keep the master branch as it is, and do rebasing on a 
separate branch. I don't know, I'm no git-master ;).

If you have no changes of your own in the branch, you can do
git reset --hard origin/master
which will reset that branch to the exact version that is in origin/master.

> 
> n.b. you earlier told me that git was better than sliced
> bread; I'm still waiting to be shown the light :-)

You'll see it when you understand how git works ;)

> 
> Second question; I've set up my system much the same as the
> boards you are working on, with an LCD and TV outputs.  I'm
> a bit confused as to how I configure the overlays and managers
> to be able to send YUV data directly through to either the
> LCD or TV.  Can you explain this?

You need to change the framebuffer to YUV color mode. And YUV only works 
on video overlays, so you need to setup video overlay to be shown on the 
LCD or TV.

If you have fb1 as a video overlay, you can do for example:

fbset -fb /dev/fb1 -nonstd 1

which sets fb1 to YUV422 mode. 8 would be YUY422. Those numbers come 
from omapfb.h, enum omapfb_color_format.

Check also Documentation/arm/OMAP/DSS, if you haven't already done so.

  Tomi

Re: DSS2 questions

[Gary Thomas]

[-- Attachment #1: Type: text/plain, Size: 2808 bytes --]

Tomi Valkeinen wrote:
> Hi,
> 
> ext Gary Thomas wrote:
>> Tomi,
>>
>> I've been following your DSS2 work, with reasonable success.
>> However, when I tried to pull from your repo today, I got
>> a ton of merge errors.  I cloned your tree a [little] while
>> back (it seems it was 2009-06-16, but where did the time go?)
>>
>> Here's what I did:
>>   git clone http://www.bat.org/~tomba/git/linux-omap-dss.git
>>     HEAD=5bd374f9b199d46ae434489d56615b240de5a6d7
>>   git checkout -b my_branch origin/master
>>     ...
>> When I went to pull into my_branch, I got the merge errors.
>> Any clues what I did wrong?  How could/should I have managed
>> this better?
> 
> The problem is that I have rebased my tree so that I end up with a clean
> set of patches that can be posted to mailing lists. And rebasing means
> changing history, which means that you cannot merge it normally.
> 
> Perhaps I should keep the master branch as it is, and do rebasing on a
> separate branch. I don't know, I'm no git-master ;).
> 
> If you have no changes of your own in the branch, you can do
> git reset --hard origin/master
> which will reset that branch to the exact version that is in origin/master.
> 

I'll give this a try - I still have the old tree I was working from.

>>
>> n.b. you earlier told me that git was better than sliced
>> bread; I'm still waiting to be shown the light :-)
> 
> You'll see it when you understand how git works ;)
> 
>>
>> Second question; I've set up my system much the same as the
>> boards you are working on, with an LCD and TV outputs.  I'm
>> a bit confused as to how I configure the overlays and managers
>> to be able to send YUV data directly through to either the
>> LCD or TV.  Can you explain this?
> 
> You need to change the framebuffer to YUV color mode. And YUV only works
> on video overlays, so you need to setup video overlay to be shown on the
> LCD or TV.
> 
> If you have fb1 as a video overlay, you can do for example:
> 
> fbset -fb /dev/fb1 -nonstd 1
> 
> which sets fb1 to YUV422 mode. 8 would be YUY422. Those numbers come
> from omapfb.h, enum omapfb_color_format.
> 
> Check also Documentation/arm/OMAP/DSS, if you haven't already done so.
> 

Still a bit fuzzy on this.  I'm sure that document helps, if
one already knows how all the overlays and managers and ...
are wired together.  I'm new at this game :-)

I tried to set my TV to be video/uyv422 via the attached script.
I was trying to leave fb0 alone (running on the LCD) and use fb1
to display the YUV422 data. Alas, 'fbset' gives me "Bad video mode"?
Any pointers?

Thanks

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

[-- Attachment #2: tv_video --]
[-- Type: text/plain, Size: 837 bytes --]

#! /bin/sh

ovl0=/sys/devices/platform/omapdss/overlay0
ovl1=/sys/devices/platform/omapdss/overlay1
ovl2=/sys/devices/platform/omapdss/overlay2

mgr0=/sys/devices/platform/omapdss/manager0
mgr1=/sys/devices/platform/omapdss/manager1

lcd=/sys/devices/platform/omapdss/display0
tv=/sys/devices/platform/omapdss/display1

fb0=/sys/class/graphics/fb0
fb1=/sys/class/graphics/fb1
fb2=/sys/class/graphics/fb2

MODE=${MODE:-ntsc}
echo "0" > $tv/enabled
echo ${MODE} >$tv/timings
w=`cat $tv/timings | cut -d "," -f 2 | cut -d "/" -f 1`
h=`cat $tv/timings | cut -d "," -f 3 | cut -d "/" -f 1`
echo "... Setting ${MODE} mode, Width = ${w}, Height = ${h}"

echo "0" > $ovl2/enabled
echo "2" > $fb1/overlays

echo "$w,$h" > $ovl2/output_size
echo "tv" > $ovl2/manager

echo "1" > $ovl2/enabled
echo "1" > $tv/enabled

fbset -fb /dev/fb1 -nonstd 1

Re: DSS2 questions

[Tomi Valkeinen]

Hi,

ext Gary Thomas wrote:
> Tomi Valkeinen wrote:
>> Hi,
>>
>> ext Gary Thomas wrote:
>>> Tomi,
>>>
>>> I've been following your DSS2 work, with reasonable success.
>>> However, when I tried to pull from your repo today, I got
>>> a ton of merge errors.  I cloned your tree a [little] while
>>> back (it seems it was 2009-06-16, but where did the time go?)
>>>
>>> Here's what I did:
>>>   git clone http://www.bat.org/~tomba/git/linux-omap-dss.git
>>>     HEAD=5bd374f9b199d46ae434489d56615b240de5a6d7
>>>   git checkout -b my_branch origin/master
>>>     ...
>>> When I went to pull into my_branch, I got the merge errors.
>>> Any clues what I did wrong?  How could/should I have managed
>>> this better?
>> The problem is that I have rebased my tree so that I end up with a clean
>> set of patches that can be posted to mailing lists. And rebasing means
>> changing history, which means that you cannot merge it normally.
>>
>> Perhaps I should keep the master branch as it is, and do rebasing on a
>> separate branch. I don't know, I'm no git-master ;).
>>
>> If you have no changes of your own in the branch, you can do
>> git reset --hard origin/master
>> which will reset that branch to the exact version that is in origin/master.
>>
> 
> I'll give this a try - I still have the old tree I was working from.
> 
>>> n.b. you earlier told me that git was better than sliced
>>> bread; I'm still waiting to be shown the light :-)
>> You'll see it when you understand how git works ;)
>>
>>> Second question; I've set up my system much the same as the
>>> boards you are working on, with an LCD and TV outputs.  I'm
>>> a bit confused as to how I configure the overlays and managers
>>> to be able to send YUV data directly through to either the
>>> LCD or TV.  Can you explain this?
>> You need to change the framebuffer to YUV color mode. And YUV only works
>> on video overlays, so you need to setup video overlay to be shown on the
>> LCD or TV.
>>
>> If you have fb1 as a video overlay, you can do for example:
>>
>> fbset -fb /dev/fb1 -nonstd 1
>>
>> which sets fb1 to YUV422 mode. 8 would be YUY422. Those numbers come
>> from omapfb.h, enum omapfb_color_format.
>>
>> Check also Documentation/arm/OMAP/DSS, if you haven't already done so.
>>
> 
> Still a bit fuzzy on this.  I'm sure that document helps, if
> one already knows how all the overlays and managers and ...
> are wired together.  I'm new at this game :-)
> 
> I tried to set my TV to be video/uyv422 via the attached script.
> I was trying to leave fb0 alone (running on the LCD) and use fb1
> to display the YUV422 data. Alas, 'fbset' gives me "Bad video mode"?
> Any pointers?
> 

You need to allocate memory for the framebuffer.

Try adding for example this:

mem=$((w*h*4))
echo $mem > $fb1/size

Also, you are trying to connect video2 overlay to fb1, but by default 
DSS connects video2 to fb2. So if you really want to do that, you first 
need to detach video1 from fb1, video2 from fb2, and then attach video2 
to fb1. I suggest you use fb2 which is already attached to video2.

  Tomi

Re: DSS2 questions

[Gary Thomas]

Tomi Valkeinen wrote:
> Hi,
> 
> ext Gary Thomas wrote:
>>>> Second question; I've set up my system much the same as the
>>>> boards you are working on, with an LCD and TV outputs.  I'm
>>>> a bit confused as to how I configure the overlays and managers
>>>> to be able to send YUV data directly through to either the
>>>> LCD or TV.  Can you explain this?
>>> You need to change the framebuffer to YUV color mode. And YUV only works
>>> on video overlays, so you need to setup video overlay to be shown on the
>>> LCD or TV.
>>>
>>> If you have fb1 as a video overlay, you can do for example:
>>>
>>> fbset -fb /dev/fb1 -nonstd 1
>>>
>>> which sets fb1 to YUV422 mode. 8 would be YUY422. Those numbers come
>>> from omapfb.h, enum omapfb_color_format.
>>>
>>> Check also Documentation/arm/OMAP/DSS, if you haven't already done so.
>>>
>>
>> Still a bit fuzzy on this.  I'm sure that document helps, if
>> one already knows how all the overlays and managers and ...
>> are wired together.  I'm new at this game :-)
>>
>> I tried to set my TV to be video/uyv422 via the attached script.
>> I was trying to leave fb0 alone (running on the LCD) and use fb1
>> to display the YUV422 data. Alas, 'fbset' gives me "Bad video mode"?
>> Any pointers?
>>
> 
> You need to allocate memory for the framebuffer.
> 
> Try adding for example this:
> 
> mem=$((w*h*4))
> echo $mem > $fb1/size
> 
> Also, you are trying to connect video2 overlay to fb1, but by default
> DSS connects video2 to fb2. So if you really want to do that, you first
> need to detach video1 from fb1, video2 from fb2, and then attach video2
> to fb1. I suggest you use fb2 which is already attached to video2.

That helped, thanks.  'fbset' no longer complains and I now have
a lovely green screen on my TV.

Now, to figure out how to send some YUV422 data to it.  I have Vaibhav's
V4L2 support in my tree - will this work with your DSS2 devices?

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------