https://bugzilla.kernel.org/show_bug.cgi?id=60824 --- Comment #82 from Swyter (swyterzone@xxxxxxxxx) --- Hmm. That's interesting, so HCI reports that the chip manufacturer is fine; Cambridge Silicon Radio (10), but then the following check fails and treats it as counterfeit... Which (if I've done my research correctly) *isn't true*, your dongle should be legit and suffering from the Stored Link Key bug, like mine, which also never showed up in `bluetoothctl` before this. ¯\_(ツ)_/¯ If you could try changing this part: > + if (le16_to_cpu(rp->manufacturer) == 10 && > + le16_to_cpu(rp->hci_rev) == le16_to_cpu(rp->lmp_subver)) { ...to this: > + if (le16_to_cpu(rp->manufacturer) == 10) { ...and find that it works for you, then we may be onto something. Until now all the CSR chips I've found put the same little version numbers in both fields, but it may not be as constant as I thought. Back to basics. In any case doing `while true; do sudo hcidump -X; done` in a terminal window that keeps running (you can stop it with Ctrl + C) even when your thingie is disconnected, then trying to plug it in, and then copying the interesting parts of that to GitHub Gist or some other paste place to link here may be useful. If you are on Arch there's a handy bluez-utils-compat AUR package with these utilities. Some interesting keywords to look for in the logs, any errors are probably useful, given some context: - Error: Unsupported Feature or Parameter Value - Stored Link Key Thanks again for testing! -- You are receiving this mail because: You are the assignee for the bug.