• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 97 subscribers
  • Views 404 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.

XDM: xdm1 buf desc for decoder output buffers and ti apps.

Omkiran4612
Prodigy 110 points

Hi,

 

We have already closed out a discussion wrt to encoders here: http://e2e.ti.com/support/embedded/f/356/p/90267/315505.aspx

Just wanted to close it out on the decoder side too.

   This is with respect to VIDDEC2 displayBufs understanding.   The displaybufs have a bufDesc which has a bufSize parameter.  My understanding is this should be the actual size of the buffer because this is the size CE uses to translate back from physical to virtual and also from the definition of xdm1_bufdesc that it is the size of the buffer.

    As output can be with a stride or with different chroma formats, the user should use a combination of pitch, width, height and chroma format in the display buffer to figure out the valid data portion.  Can you confirm if this understanding is true?

 The reason I ask this is because TI codec apps on 3530 seem to use the bufSize directly to dump the buffer.  Hence there is a confusion on what bufSize should be [valid data in output or actual buffer size].  Or is it just to simplify the application?

Regards,
Omkiran

 

  • 0
    TI__Genius 17205 points

    _My_ understanding is that _your_ understanding is correct.  :)  The IVIDDEC2_OutArgs.displayBufs.bufDesc[].bufSize field indicates the size of the actual memory buffer, and doesn't indicate anything about the contents of the buffer.  The other fields you've listed can be used to interpret the contents of that buffer.

    The next conclusion to draw, I guess, is that the apps you refer to are written incorrectly...

    Assuming no one disputes my claim, I'll plan to update the XDM wording to clarify this - similar to what we did for IVIDENC1.

    Chris