https://bugzilla.kernel.org/show_bug.cgi?id=207629 --- Comment #21 from Abhishek Pandit-Subedi (abhishekpandit@xxxxxxxxxxxx) --- re #19 - Chromium had an internal patch that was conditionally sending the passive scanning command. After removing it, we were getting command disallowed. This is very likely what you were seeing. There's a patch fixing this on the mailing list now: https://patchwork.kernel.org/patch/11582831/ re #20 - You make a good point. A second problem I encountered with higher usage of this code is that if the controller stops responding (i.e. it crashes) and we can't reset it, it will also prevent suspend. This one is unacceptable. Likely, the fix from Bluetooth end would be just to always allow suspend but log an error here: https://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next.git/tree/net/bluetooth/hci_core.c#n3368 Expect a patch today on the mailing list with a fix for this. -- You are receiving this mail because: You are the assignee for the bug.
