MT7925 Bluetooth fails on Thelio (0e8d:0717) — WMT command timeout due to SLPPROT

[messinadm]

System: Pop!_OS 24.04, Cosmic, kernel 6.18.7-76061807-generic
Hardware: System76 Thelio, MT7925 PCIe combo card (USB ID 0e8d:0717)

Symptom: Bluetooth cannot be enabled. btmgmt info shows 0 controllers.

dmesg:
Bluetooth: hci0: Execution of wmt command timed out
Bluetooth: hci0: Failed to send wmt patch dwnld (-110)
Bluetooth: hci0: Failed to set up firmware (-110)

Root cause: The MT7925 BT core's sleep protection register (0x18011100
bit 1) keeps the BT processor asleep during driver initialization. WMT
commands reach the device (USB Control OUT status=0) but the device never
responds. Additionally, after firmware activation the chip re-enables
sleep protection, blocking the subsequent FUNC_CTRL command.

Fix: In btmtk_usb_setup() for dev_id 0x7925:

1. Call btmtk_usb_subsys_reset() to clear stuck WMT state
2. Set 0x18011100 |= BIT(1) before firmware download
3. Set 0x18011100 |= BIT(1) again before FUNC_CTRL

Status: Patch submitted upstream to linux-bluetooth@vger.kernel.org
on 2026-05-31: "[PATCH] Bluetooth: btmtk: Fix MT7925 WMT command timeout
due to sleep protection"

Requesting backport to the System76 kernel builds.

[messinadm]

Update: the upstream fix has landed and is confirmed working.

A MediaTek engineer (Chris Lu) responded to the linux-bluetooth thread and confirmed their fix for this issue shipped in kernel 7.1-rc. I tested on 7.1.0-0.rc6.260604gba3e43a9e601.43.fc45.x86_64 — Bluetooth initializes cleanly with no WMT timeouts, and a Bluetooth HID device connected successfully.

Requesting System76 backport Chris Lu's btmtk reset/firmware initialization fix from kernel 7.1-rc to the 6.18.x series. The patch series is titled "Bluetooth: btmtk: Add reset mechanism if downloading firmware failed" and was merged into bluetooth-next around February 2026.

My own patch (submitted May 31 to linux-bluetooth@vger.kernel.org, subject "Bluetooth: btmtk: Fix MT7925 WMT command timeout due to sleep protection") addresses the same root cause and works on 6.18.7 in the meantime if the upstream backport is not feasible.

[mmstick]

The 6.18 kernel is no more. 7.0.11 will be released this week.

[messinadm]

The fix landed in 7.1-rc. ill update and reopen this issue if my mira can't do Bluetooth

[armilancode]

I’m hitting what looks like a related MediaTek btmtk Bluetooth regression on Pop!_OS 24.04, but on MT7922 rather than MT7925.

System:

• Pop!_OS 24.04 LTS
• Broken kernel: 7.0.9-76070009-generic
• Previously working kernel: 6.18.7-76061807-generic
• PCI Wi-Fi device: MediaTek MT7922 [14c3:0616]
• USB Bluetooth device: 0489:e0f6 MediaTek Inc. Wireless_Device

Symptoms:

• bluetooth.service is running
• rfkill shows Bluetooth is not blocked
• bluetoothctl reports: No default controller available
• hci0 exists but is not initialized

Relevant broken-kernel log:

Bluetooth: hci0: HW/SW Version: 0x008a008a, Build Time: 20260106153735
Bluetooth: hci0: Failed to send wmt func ctrl (-22)
Bluetooth: hci0: HCI Enhanced Setup Synchronous Connection command is advertised, but not supported.

hciconfig -a on the broken kernel:

hci0: Type: Primary  Bus: USB
    BD Address: 00:00:00:00:00:00  ACL MTU: 0:0  SCO MTU: 0:0
    DOWN
    Features: 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00

The previous 6.18.7 Pop kernel initialized the same adapter correctly:

Bluetooth: hci0: Device setup in 153166 usecs
Bluetooth: hci0: AOSP extensions version v1.00
Bluetooth: hci0: AOSP quality report is supported

I’ll test again when Pop kernel 7.0.11 is published. If 7.0.11 is expected to contain the 7.1-rc btmtk fixes, this should be a useful confirmation case for MT7922 / 0489:e0f6 as well.