theora dev - Jul 2007 - Theora Hardware: Integration with LEON is completed!

Hi,

The Integration with LEON (first part of my GSoC) is completed.
http://atlas.lsc.ic.unicamp.br/~andre.lnc/theora_integration_with_leon3_full.png

At the last week I had (leon3 + linux + libtheora) and  (leon3 +
send_vector_of_input + theora hardware) working ok.
Firstly, I thougth that it just would be:
http://atlas.lsc.ic.unicamp.br/~andre.lnc/theora_integration_with_leon3.png
But it wasn't too simple, when I joined this I had a lot of problems with
linux, because I discovered that I need to do a Driver (using linux kernal
API) that works with leon3/linux- 2.6 and commads our Theora hardware.

I didn't have any idea about drive and I spend these last 4 day learning how
to do a drive for my specific device. The sparc_linux email list helped me a
lot and now I solved all problems (mainly about MMU).

* What I already know to do (First part of GSoC2007 - Completed)
 - To study LEON3 and to find a configuration that be able to decode the
Theora and to do a synthesis of LEON3 (OK - Completed)
 - To compile the Theora for LEON and to run this inside of the LEON/FPGA
(OK - Completed)
 - To study and to discovery how to plug the Theora Hardware on AMBA and to
control it with the LEON  (OK - Completed)
 - To change the handshake of Theora Hardware of AVALON to AMBA protocol (OK
- Completed)
 - Running Theora decoder on LEON (FPGA) decoding a vetor of inputs (OK -
Completed)
 - To study and to implement and Linux Device Driver for Theora Hardware (OK
- Completed)
 - To join the software interface with Theora hardware on linux and LEON3
(OK - Completed)

Now, I was thinking to do a webpage explaining all the step that are needed
to do and to submit all that files changed and added.

* Next Step of my work on GSoC
I think next step would be the to do a video controler. Then, we will have
all the process on FPGA at final of my GSoC!
I just was having a doubt. Is the UpdateUMV the last step of decodification?


I saw on libtheora that after UpdateUMV there is the function PostProcess.
void PostProcess(PB_INSTANCE *pbi)
Is this function just a optimization for software? (that would be unecessary
for hardware)

Do you think that will be necessary more one module of decodification
between UpdateUMV and video controler?

Cheers,

-- 
Andr? Costa

Gerente T?cnico
Projeto BrazilIP
LSC IC-UNICAMP

Cel: + 55 13 9201 1870
http://www.brazilip.org.br/
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/theora-dev/attachments/20070711/9ece02fd/attachment.htm