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.

Linux/AM3352: slow to plug-in of USB hub after communication Freeze patch.

Part Number: AM3352

Tool/software: Linux

Dear Champs,

My customer is using Micron's USB hub to connect 2ea of thermal cameras, and resolved hang issue by using below patches.

https://e2e.ti.com/support/processors/f/791/p/792039/2938895?tisearch=e2e-sitesearch&keymatch=usb%20AND%20patch#2938895

But, after patch, and when there was sudden unplugged of USB hub occurred, it was too slow to be plug-in again of these thermal cameras.

Before patch, it takes only 6 secs to plug-in thermal cameras, but now it takes almost 13 secs after patch.

Is there any other USB patch to help this issue?

Thanks and Best Regards,

SI.

  • Please post which Linux version is used.
  • Linux Processor SDK v4.01.

    Thanks and Best Regards,
    SI.
  • Hi,

    Is the hub a high-speed hub? Are the thermal camera are usb full-speed?
    Is the AM3352 usb port configured as host or otg mode?
    Can you please use a scope to probe USB_VBUS and the hub D+ signals to see how both transitions during the 13 secods?
  • Hi,

    Yes. their USB Hug is Micronchip's USB2512B-I/M2 supporting high speed, but they connected thermal camera through usb with full speed.
    In USB1 of AM3352, this USB hub was connected, and these 2 thermal cameras were connected to this USB hub through USB full speed.
    In normal case, these thermal camera uses V4L2 to capture image through /dev/video0 and /dev/video1 at each.

    Unfortunately they could not measure D+ signal yet, but they found below error occurred in hub.c when slow connection issue was occurred
    clear tt (8xxx) error -110

    Could you please let me know what this error mean?
    This error was occurred at every time when issue occurred - the connection was to be slow. In normal case, this error was not occurred.


    Thanks and Best Regards,
    SI.

  • Hi Sung-IL,

    I am not sure if it is related, but you can see from the comment in the link below, the patch which fixes the clearing TT buffer is not complete. But I don't have a setup to reproduce the hub lockup issue so am not sure how to improve the patch.

    www.spinics.net/.../msg162310.html