[Bug 217475] New: Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[Bug 217475] New: Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

            Bug ID: 217475
           Summary: Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops
                    working on ThinkPad T14s AMD Gen1
           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: berto@igalia.com
        Regression: No

Hi,

I'm running Linux on a ThinkPad T14s AMD Gen 1 and I'm having problems
with its Bluetooth adapter.

The adapter is a Foxconn / Hon Hai, USB ID 0489:e0cd, and it uses the
Mediatek MT7921 driver. Sometimes it just works but often if I suspend
the laptop and come back it stops working. This also happens when I
use the RF kill switch. The kernel errors vary a bit depending on the
moment but it's usually a combination of these:

kernel: Bluetooth: hci0: Failed to get device id (-108)
kernel: Bluetooth: hci0: Failed to get fw version (-108)
kernel: bluetooth hci0: firmware: direct-loading firmware
mediatek/BT_RAM_CODE_MT7961_1_2_hdr.bin
kernel: Bluetooth: hci0: Execution of wmt command timed out
kernel: Bluetooth: hci0: Failed to send wmt patch dwnld (-110)
kernel: Bluetooth: hci0: Failed to set up firmware (-110)
kernel: Bluetooth: hci0: HCI Enhanced Setup Synchronous Connection command is
advertised, but not supported.

Interestingly I noticed that if I try to use it on a Linux VM (using
USB passthrough) it generally works better, even when it's not working
on the host and the guest and the host are running the same OS and
kernel version. Still it fails sometimes, although the error messages
tend to be a bit different:

kernel: usbcore: registered new interface driver btusb
kernel: Bluetooth: hci0: Device setup in 153593 usecs
kernel: Bluetooth: hci0: HCI Enhanced Setup Synchronous Connection command is
advertised, but not supported.
kernel: Bluetooth: hci0: Opcode 0x c03 failed: -110
kernel: Bluetooth: hci0: Failed to read MSFT supported features (-110)
kernel: Bluetooth: hci0: AOSP get vendor capabilities (-110)

(I also tried the adapter with a Windows VM, again using USB
passthrough, but here it works perfectly fine)

Back to the host I found out that resetting the adapter using
"usb_modeswitch -R -v 0489 -p e0cd" sometimes is enough to bring the
adapter back to life and make it work. But this doesn't always
succeed, and at some point I end up in a situation like this:

kernel: xhci_hcd 0000:06:00.4: xHCI host not responding to stop endpoint
command
kernel: xhci_hcd 0000:06:00.4: xHCI host controller not responding, assume dead
kernel: xhci_hcd 0000:06:00.4: HC died; cleaning up

Sometimes I can get out of it with this:

$ echo 0000:06:00.4 > /sys/bus/pci/drivers/xhci_hcd/unbind
$ echo 0000:06:00.4 > /sys/bus/pci/drivers/xhci_hcd/bind

All this has been happening since I got the laptop, but I'm testing it
now with Linux 6.3.3 and the problem is still there.

The end result is that I basically cannot rely on the USB adapter so I
have it almost always disabled.

A bit more information about the system:

$ lspci -nn | grep USB
02:00.4 USB controller [0c03]: Realtek Semiconductor Co., Ltd. RTL811x EHCI
host controller [10ec:816d] (rev 0e)
05:00.0 USB controller [0c03]: Renesas Technology Corp. uPD720202 USB 3.0 Host
Controller [1912:0015] (rev 02)
06:00.3 USB controller [0c03]: Advanced Micro Devices, Inc. [AMD]
Renoir/Cezanne USB 3.1 [1022:1639]
06:00.4 USB controller [0c03]: Advanced Micro Devices, Inc. [AMD]
Renoir/Cezanne USB 3.1 [1022:1639]

$ lsusb
Bus 007 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 006 Device 003: ID 0489:e0cd Foxconn / Hon Hai Wireless_Device
Bus 006 Device 002: ID 06cb:00bd Synaptics, Inc. Prometheus MIS Touch
Fingerprint Reader
Bus 006 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 005 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 004 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 003 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 002 Device 002: ID 04f2:b6cb Chicony Electronics Co., Ltd Integrated Camera
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

Bagas Sanjaya (bagasdotme@gmail.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bagasdotme@gmail.com

--- Comment #1 from Bagas Sanjaya (bagasdotme@gmail.com) ---
What kernel version did you have this issue? If it is regression, what is last
working version?

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

--- Comment #2 from Alberto Garcia (berto@igalia.com) ---
I have always had this problem with this computer. Linux got support for this
Bluetooth adapter in 5.14 if I'm not wrong (64832df2ac05) and I probably tried
it for the first time with 5.15.3. I never saw it working fine, and I have
tried lots of kernels and firmware versions since then. The most recent one was
6.3.3 as I mentioned in the bug report.

Since I can make it work with a Windows VM using USB passthrough I assume that
there's some problem with the initialization of the adapter, the HCI or
something like that, but that's all I can guess with my limited knowledge of
USB.

To make it clear: when I talk about using a VM I'm talking about the scenario
where

1) I cannot use Bluetooth in my Linux system because of one of the errors that
I mentioned earlier ("hci0: Failed to get device id (-108)"). The device is not
detected by the Bluetooth stack although it does appears in lsusb.
2) I launch a Windows VM using QEMU right away, without rebooting the computer
or doing anything else. The BT adapter works in that VM always.
3) If I launch a Linux VM instead it tends to work, not always in this case but
I have a higher success rate than if I remove and reload the btusb module on
the host.

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

--- Comment #3 from Alberto Garcia (berto@igalia.com) ---
One more thing: when it's working on Linux it seems to be reliable, at least I
have never had the situation where I'm listening to music and it suddenly stops
working.

In my experience the problem appears when the adapter is disabled and then
re-enabled, either when the laptop is suspended or with the RF kill switch.

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

Leonardo Silveira (sombriks@gmail.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sombriks@gmail.com

--- Comment #4 from Leonardo Silveira (sombriks@gmail.com) ---
i got similar issue on ThinkPad T14 Gen 2a

i turned it off, slapped the machine twice then it went back working.
not completely sure if it's a software issue but it is happening.

when it stops working it keeps not working. until it come back.

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

--- Comment #5 from Alberto Garcia (berto@igalia.com) ---
Hi, I'm on Linux 6.12.73 now.

I found that btusb.enable_autosuspend=n seems to solve the problem.

Commit 67dba2c28fe0 ("btmtk: Fix failed to send func ctrl for MediaTek
devices.") prevents the device from suspending during btmtk_usb_shutdown() to
avoid timeout errors similar to the ones I'm experiencing:

   btmtk_usb_hci_wmt_sync() hci0: Execution of wmt command timed out
   btmtk_usb_shutdown() hci0: Failed to send wmt func ctrl (-110)

So I wonder if the same thing is happening during btmtk_usb_setup().

There's a long gap until the bluetooth driver is loaded. This computer has an
external USB-C dock that sometimes takes a long time to initialize, also the
hard drive is encrypted with LUKS so the btusb driver (and its firmware) cannot
be loaded until I enter the password and the rootfs is decrypted. All this can
take easily 20 seconds.

So maybe the bluetooth adapter suspends before the driver is loaded and that's
why the initialization fails, does that make sense?

I tried patching btmtk_usb_setup() with the same approach and so far I haven't
been able to reproduce the issue anymore.

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

Alberto Garcia (berto@igalia.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |berto@igalia.com

--- Comment #6 from Alberto Garcia (berto@igalia.com) ---
Created attachment 309604
  --> https://bugzilla.kernel.org/attachment.cgi?id=309604&action=edit
Proposed patch: add USB autopm support to btmtk_usb_setup()

Here's the patch that I'm using, I haven't been able to reproduce the bug after
several reboots and suspend / resume cycles.

-- 
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 217475] Foxconn / Hon Hai Bluetooth adapter 0489:e0cd stops working on ThinkPad T14s AMD Gen1

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=217475

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

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

--- Comment #7 from Michał Pecio (michal.pecio@gmail.com) ---
Does it mean that with this patch you no longer need to disable autosuspend and
all those problems are gone?

It makes some sense, as it seems that btusb_probe() only stores a pointer to
btusb_mtk_setup() and then enables autosuspend, so somebody else may later call
that function on a suspendend device. Other BT drivers may be affected too.

If you got this far, please email your findings to bluetooth maintainers
responsible for this driver (run ./scripts/get_maintainer on the buggy file).
If they find your solution acceptable you may be asked to email the patch too.

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

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