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: ldc report tivxVpacLdcProcess error

Part Number: TDA4VM

Tool/software:

Hi Ti Experts:

    We need to manually obtain the output of ldc through vx_GraphParameterDequeueDoneRef for processing. Then, following the demo, we configure as shown in the figure below and obtain the image.

However, we found that an error message "[tivxVpacLdcProcess:323] Failed to Submit Request" was printed when printing 32 lines (4-channel camera images with a buffer depth of 8). Then we located the error occurrence at the place shown in the following screenshot.

We also discovered that if we called vxMapImagePatch before vx_GraphParameterDequeueDoneRef to allocate space for the array created, there would be no error and no error message would be printed after printing 32 lines. This is because we called this function. My question is, is this calling method correct? Or what should the correct process be like?We are using SDK version 8.4 on a Linux platform.

Thanks and best regards.

  • Hi Jia,

    In the code snippet you sent,

    You shouldn't be assigning intermediate_obj_arr to ldcobj.output_arr after creating ldc graph, because ldc node is created with elements from ldcobj.output_arr, check that once.

    If your goal is to save the ldc output image to a file then I would recommend to use the en_out_img_write configuration used in the multi_cam_app

    Regards,
    Gokul

  • Hi,

    I'm glad to receive your reply. From multi_cam_app, I understand that en_out_img_write is used to store the final mosaic image. However, what I need is to obtain the image of ldc for our subsequent processing,The output of ldc is not utilized by other OpenVX nodes.Although you mentioned above that our configuration is incorrect, we can obtain the image normally after using vxMapImagePatch. May I ask how to obtain it correctly?

    Thanks and best regards.

  • Hi,

    I understand that en_out_img_write is used to store the final mosaic image.

    Yes, similarly you can use en_out_ldc_write to store ldc output to file.

    You shouldn't be assigning intermediate_obj_arr to ldcobj.output_arr after creating ldc graph, because ldc node is created with elements from ldcobj.output_arr,

    what i meant to say is, you are assigning ldcobj.output_arr after creating graph, try assigning before creating graph, for example,

    obj->ldcobj.output_arr = obj->intermediate_obj_arr[0];
    status = app_create_graph_ldc(obj->graph. &obj->ldcobj, ldc_in_arr);
    printf("LDC_CONVERT graph done!\n");

    Regards,
    Gokul

  • Hi,

    Yes, we have tried it this way before, but the problem still occurs.

    Thanks and best regards.

  • Hi,

    Is this the same issue that we are discussing on the other thread? 

    Regards,

    Brijesh

  • Hi,

    Please refer to the below patch for app_multi_cam to make ldc output as graph parameter,

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/791/5355.ldc_5F00_graph_5F00_parameter.patch

    Regards,
    Gokul

  • Hi,

    We have set it up in this way before, but still got printing errors. However, we have identified the cause of the problem. Thank you for your reply.

    Thanks and best regards.

  • Thanks, closing this thread.