Return-Path: Date: Wed, 19 Sep 2018 09:18:49 -0300 From: Luiz Carlos Ramos To: Marcel Holtmann , Johan Hedberg Cc: linux-bluetooth@vger.kernel.org Subject: [PINGBACK] Patches for ath3k with possible issues Message-ID: <20180919121849.GA29863@giustizia.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 List-ID: Hello, I sent these messages some time ago and as I am not a frequent patch developer and hence not used to send patches to the community, I'd like to know if you have been aware of them. I haven't received any replies so far. The patches were sent in sequence of this "heading" message and are trivial, however, very useful for my case. Only to remember, they print messages in case of errors in two specific operations in the ath3k kernel module. Many thanks, Luiz Carlos Ramos lramos dot prof at yahoo dot com dot br São Paulo - Brazil ----- Forwarded message from Luiz Carlos Ramos ----- Date: Fri, 7 Sep 2018 20:35:31 -0300 From: Luiz Carlos Ramos To: Marcel Holtmann , Johan Hedberg Cc: linux-bluetooth@vger.kernel.org Subject: Patches for ath3k with possible issues User-Agent: Mutt/1.10.1 (2018-07-13) Hello, I think I have found some smaller problems related to a specific Bluetooth adapter that isn't working fine in a laptop that is with me from some time. In fact, only now I've found enough time to dig into the code and try to make some experiments. The Bluetooth adapter is an Atheros, identified by VID/PID 0cf3:0036, running in a x86_64 laptop. The problems are just like that: 1- when the laptop starts, Bluetooth always work; however, I found that the "switch vid/pid" operation doesn't work, returning -32 ("EPIPE") 2- after a suspend or hibernate, when it resumes, sometimes the firmware download doesn't work. It stops before sending all pieces and when it happens, it returns code -32 ("EPIPE") 3- there are cases also where after a resume, the switch VID/PID works and the adapter changes its identification to 0cf3:3007 4- in cases 2 and 3, sometimes the bluetooth daemon doesn't recognize the adapter anymore. In my tests, I can restore operation doing in sequence: rmmod ath3k, insmod ath3k, kill bluetooth daemon and start bluetooth daemon again Even not knowing the problem in detail right now, I can send some patches which helped me to collect the information above, specially the return code of the switch VID/PID operation and the state of the firmware load when the error happens. I'll send you next two patches which addresses these two things, and hopefully this can help someone else that stays in trouble with problems like those. Many thanks, Luiz Carlos Ramos lramos dot prof at yahoo dot com dot br São Paulo - Brazil ----- End forwarded message -----