• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 102 subscribers
  • Views 905 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DM368 VPFE ISP H3A AEWB failure issue

frank lin56
Prodigy 30 points

Hello

We got an issue: DM368 H3A AEWB statistics result become abnormal and cannot be recovered.  We think the issue happens because of temporarily unstable Hsync from sensor.  Could you help us to solve this issue: 

1. How can we check the status of H3A and detect the failure of H3A? 

2. This issue cannot be recovered until chip HW reset.  Is there any way to recover this issue by firmware resest?

Thanks

  • 0
    Guru 11260 points

    Hi,

    How can we check the status of H3A and detect the failure of H3A?

    By failure of H3A what exctly do you mean ? does it crash or something ?

    When you detect the unstable HSYNC from sensor why not restart the H3A have you tried it ? (by restart I mean close it and open again /configure start)

    Thanks,

    --Prabhakar Lad

  • 0 in reply to Prabhakar Lad
    Prodigy 30 points

    Hi,

              We judge that H3A AEWB failed because the statistics results from H3A AEWB is not changed even if image data from sensor changed a lot.   

              Till now, we cannot detect the unstable HSYNC. Thus, we want to know the method to detect this failure.  

               Also, after this failure appears, we try to disable and then enable AEWB by setting register H3A_PCR[16] AEW_EN bits. But this failure cannot be recovered.

    Thanks

             

     

  • 0 in reply to frank lin56
    Prodigy 195 points

    The H3A module does not cope very well with unstable sync signals. Not only can the statistics stop updating, but also the H3A hardware block collection can "run away" continuing to collect after the end of the buffer in memory, stomping on unrelated memory. So it's vital that you fix that unstable HSYNC signal.

    Meanwhile you can put in some basic detection like magic numbers in a few locations in the H3A buffer so that you can tell if they have been modified after collection, preferably using bit combinations for the magic numbers that are impossible to appear in the real H3A stats. If you detect an error you could trying resetting the VPSS using DRV_SyncRst(). However if it has corrupted memory elsewhere then it is too late, so really you need to fix that unstable sync if you want to use H3A.

    Kelvin.

  • 0 in reply to KAL1
    Prodigy 30 points

    Dear Kelvin

    Thanks for your reply.  We will try it using your suggestion.