USB HDD xhci_drop_endpoint

1st Official Post

    Hello,

    this is a bit of a follow-up to previous question.

    I have Beelink GT1 (S912) with external drive USB HDD WD My Passport 25E2 (4TB) and when I connected another WD My Passport (2TB) I started getting I/O error (more see the previous question).

    I though maybe the GT1 does not have enough power for both drives so I got USB hub with external power adapter.

    Before I could finish the previous question I wanted to backup my data and got a new drive - Seagate M3 Portable (4TB).

    When I connected the WD 4TB and Seagate M3 (or Maxtor M3) and let it run over night I got dmesg full of this xhci_drop_endpoint and eventually the drive stopped responding - would hang when trying 'ls' the mounted partition.

    Code
    [  131.968297@7] usb 1-1.2: reset high-speed USB device number 4 using xhci-hcd
[  131.968341@7] xhci-hcd xhci-hcd.0.auto: Setup ERROR: setup context command for slot 3.
[  132.048257@7] usb 1-1.2: reset high-speed USB device number 4 using xhci-hcd
[  132.048298@7] xhci-hcd xhci-hcd.0.auto: Setup ERROR: setup context command for slot 3.
[  132.128261@7] usb 1-1.2: reset high-speed USB device number 4 using xhci-hcd
[  132.168911@4] xhci-hcd xhci-hcd.0.auto: xHCI xhci_drop_endpoint called with disabled ep ffffffc0620ecd80
[  132.168929@4] xhci-hcd xhci-hcd.0.auto: xHCI xhci_drop_endpoint called with disabled ep ffffffc0620ecdc0
[  147.791177@3] xhci-hcd xhci-hcd.0.auto: WARN Event TRB for slot 2 ep 2 with no TDs queued?


    Full log but only with few minutes of runtime, because the overnight log.got cut.

    Dmesg part of the overnight log - dmesg

    • Official Post

    Sadly "user reports bad behaviour from Amlogic legacy kernel sources" is not surprising or something anyone will investigate. We have abandoned that kernel (which is full of bad, period) in favour of the upstream/mainline kernel which is maintainable, although that work is still ongoing and current test images you can find on Index of / are not fully featured. It's a bit of a "rock and a hard place" situation, sorry.

    Quote from chewitt

    Sadly "user reports bad behaviour from Amlogic legacy kernel sources" is not surprising or something anyone will investigate. We have abandoned that kernel (which is full of bad, period) in favour of the upstream/mainline kernel which is maintainable, although that work is still ongoing and current test images you can find on Index of / are not fully featured. It's a bit of a "rock and a hard place" situation, sorry.

    Thanks for that reply!

    I was looking for new kszaq's releases but I couldn't find any newer than I am running, so thanks for the link.

    And baring that it might not work - which image is intended for Amlogic S912 with 1Gb Ethernet and 2GB RAM?

    I assume this one LibreELEC-AMLGX.arm-9.80-nightly-20200103-587404e-box.img.gz

    or are the images more generic and not necessary to chose the right 'device tree'?

    Or even better is there a page which describes the different version?

    Small update:

    After connecting the WD 4TB through the usb hub and the Seagate (Maxtor) M3 directly to the box, the error messages stopped and so far it looks like the drive is working correctly.

    It looks like the WD 4TB doesn't care how it is connected, but the other drives are more picky.