Symptom: cheap USB Bluetooth dongles sold as "CSR 4.0/5.0/5.1/5.3 adapter" (
lsusb: 0a12:0001 Cambridge Silicon Radio) fail to pair, time out on HCI commands, or hci0 gets stuck until unplug/replug. Kernel log shows:Bluetooth: hci0: CSR: Unbranded CSR clone detected; adding workarounds...
Bluetooth: hci0: command 0x0401 tx timeout
Bluetooth: hci0: CSR: Couldn't suspend the device for our Barrot 8041a02 receive-issue workaround
Root cause: these are clone chips (Barrot 8041a02 and similar), not genuine CSR8510 hardware. They respond to some HCI init commands with malformed/incomplete data, and the kernel's existing clone-detection workaround doesn't fully compensate for it — so init either fails outright or the controller gets wedged after a timeout, with no automatic recovery.
Fix: patched
btusb to handle these malformed responses correctly during init, fixed a fragile USB power-management suspend path that made things worse, and added automatic recovery (USB reset) on init failure/timeout instead of leaving hci0 dead. Only affects the detected clone code path — genuine CSR hardware is untouched.Result: adapter initializes and pairs normally instead of needing repeated unplug/replug/reset cycles.
Caveat: this covers the specific failure modes I could reproduce and test against; different clone batches use different chips internally (as pointed out in discussions elsewhere), so it may not cover every single fake CSR8510 variant out there.
Repo (patch, DKMS install/rollback instructions): https://github.com/hhsnake/csr8510-fix
Install:
sudo apt install dkms linux-headers-$(uname -r)
git clone https://github.com/hhsnake/csr8510-fix.git
cd csr8510-fix
sudo ./install.sh
Tested on kernels 5.15–7.0+ across several Ubuntu LTS/HWE combinations.
https://redd.it/1v0anuh
@r_linux
