• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 98 subscribers
  • Views 244 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA4VM: Multiple application writing to the same display node

David Khouya
Prodigy 110 points
Part Number: TDA4VM

Hi,

I am trying to build an application which would be a display overlay over and other application. In our current configuration, the first application is using Opengl and a display node that runs in the R5F. I would like to know, is it possible to make a second application which will overlay the first application using the display node? 

Also, if using the display node, is there a configuration where we could be able to detect mouse click for clicking a button as an example?

Setup information

  • TDA4VM
  • Linux RTOS (using vision_apps dtbo)
  • SDK8.6 (Soon SDK 9.0)

Thank you,

  • 0
    TI__Guru* 84906 points

    Hi,

    Are both the displays being used runs of R5?

    Are the pipeids different in each applicaiton?

    David Khouya said:
    Also, if using the display node, is there a configuration where we could be able to detect mouse click for clicking a button as an example?

    Sorry, i did not understand this usecase, could you elaborate?

    Regards,

    Nikhil

  • 0 in reply to Nikhil Dasan
    Prodigy 110 points

    Absolutely.

    Application A is running video plaback on the R5.

    Application B is running an GUI application that I would like to overlay over the application A. As an example it could be buttons that can be clickable.

    Those are 2 different binaries, and there's no IPC between application A and application B.

    Is that a valid scenario if we are using the R5 display node or we absolutely need to fallback on linux display management. 

    Let me know if it's clearer.

  • 0 in reply to David Khouya
    TI__Guru* 84906 points

    Hi,

    2 applications could use display nodes and select separate video pipes and get the overlay output on the display.

    However, the curser part is something that Qt would do, and this is currently not supported by the display node.

    Currently, Qt is supported only for the linux display driver.

    Regards,

    Nikhil

  • 0 in reply to Nikhil Dasan
    Prodigy 110 points

    Hi,

    Do you have an example of using 2 diplay nodes in separate video pipes? Is there special configuration that need to be done for that?

    So for the cursor, we can only use it using the linux display driver. To your knowledge would it be a way to do it with the display node even if it's not currently supported from the TI's sdk?

  • +1 in reply to David Khouya
    TI__Guru* 84906 points

    Hi,

    The vision_apps utility grpx that is used along with the multicam demo could be an example for 2 display nodes in different graph.

    This utility is used to display the perf stats (i.e. core loading bars etc ) along with the video output.

    Here the multicam demo uses display node and the grpx also uses display node (but both uses different display pipes).

    The utility can be found in vision_apps/utils/grpx folder

    David Khouya said:
    So for the cursor, we can only use it using the linux display driver. To your knowledge would it be a way to do it with the display node even if it's not currently supported from the TI's sdk?

    On the linux end, this is achieved based on the Qt framework and the framebuffer is allocated from the linux display driver which also supports various IOCTLs required for Qt.

    This implementation is currently missing in the R5 driver and could be a challenge to map the framebuffer to the TIOVX framework and pass it to the display node.

    Regards,

    Nikhil

  • 0 in reply to Nikhil Dasan
    Prodigy 110 points

    That was the answer I was searching for.

    Thanks for the support Nikhil!