[Bug 219924] New: xhci_hcd TRB DMA error and device lockup with AKLOT 2i2 USB audio interface

[Bug 219924] New: xhci_hcd TRB DMA error and device lockup with AKLOT 2i2 USB audio interface

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=219924

            Bug ID: 219924
           Summary: xhci_hcd TRB DMA error and device lockup with AKLOT
                    2i2 USB audio interface
           Product: Drivers
           Version: 2.5
          Hardware: All
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P3
         Component: USB
          Assignee: drivers_usb@kernel-bugs.kernel.org
          Reporter: kernel@deimos.ca
        Regression: No

Created attachment 307892
  --> https://bugzilla.kernel.org/attachment.cgi?id=307892&action=edit
'dmesg' output while plugging device through a hub, lsusb -t, uname -a and
lspci | grep -i usb

My USB audio interface (AKLOT 2i2, Vendor 152a, Product 887e) causes a
reproducible failure in the Linux USB subsystem when unplugged during or after
audio activity. The issue is severe enough that the xHCI controller refuses to
re-enumerate the device until a full reboot (or cold plug with hub reset). My
laptop is a Lenovo Slim 7, 

In some cases, the AKLOT interface locks up internally, requiring power-cycle
or connection to another host to recover.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 219924] xhci_hcd TRB DMA error and device lockup with AKLOT 2i2 USB audio interface

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=219924

Michał Pecio (michal.pecio@gmail.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |michal.pecio@gmail.com

--- Comment #1 from Michał Pecio (michal.pecio@gmail.com) ---
> [Mar25 21:35] usb 3-5: USB disconnect, device number 7
> [  +0.000005] usb 3-5.1: USB disconnect, device number 8
> [  +0.001183] xhci_hcd 0000:00:14.0: ERROR Transfer event TRB DMA ptr not
> part of current TD ep_index 1 comp_code 1
> [  +0.000006] xhci_hcd 0000:00:14.0: Looking for event-dma 00000000ffa87930
> trb-start 00000000ffa87940 trb-end 00000000ffa87940 seg-start
> 00000000ffa87000 seg-end 00000000ffa87ff0
> [  +0.000153] xhci_hcd 0000:00:14.0: ERROR Transfer event TRB DMA ptr not
> part of current TD ep_index 1 comp_code 1
> [  +0.000001] xhci_hcd 0000:00:14.0: Looking for event-dma 00000000ffa87940
> trb-start 00000000ffa87950 trb-end 00000000ffa87950 seg-start
> 00000000ffa87000 seg-end 00000000ffa87ff0
> [  +0.000119] xhci_hcd 0000:00:14.0: ERROR Transfer event TRB DMA ptr not
> part of current TD ep_index 1 comp_code 1
> [  +0.000002] xhci_hcd 0000:00:14.0: Looking for event-dma 00000000ffa87950
> trb-start 00000000ffa87960 trb-end 00000000ffa87960 seg-start
> 00000000ffa87000 seg-end 00000000ffa87ff0

This looks like we get transfer events normally each microframe but for some
reason TDs are consistently dropped from the driver's internal list one step
too early. Very bizarre.

Are you able to test a newer kernel like 6.14 or 6.13.x, or at least the 6.12
LTS series? The 6.8 series is frankly no longer supported here, and TBH we
can't be sure if it's not a bug unique to Ubuntu 6.8 kernel due to a botched
backport of some changes from newer releases.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 219924] xhci_hcd TRB DMA error and device lockup with AKLOT 2i2 USB audio interface

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=219924

--- Comment #2 from Michał Pecio (michal.pecio@gmail.com) ---
Either that, or it's possibly related to
https://bugzilla.kernel.org/show_bug.cgi?id=218695

where IIRC some Intel controller was executing (or at least signaling
completion of) isochronous OUT transfers twice per transfer after device
disconnection.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.