Return-Path: Date: Mon, 20 Sep 2010 21:49:43 +0300 From: Johan Hedberg To: Pacho Ramos Cc: linux-bluetooth@vger.kernel.org Subject: Re: >net-wireless/bluez-4.63 unable to connect audio streams due commit Message-ID: <20100920184943.GA22968@jh-x301> References: <1285006436.15378.2.camel@localhost.localdomain> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <1285006436.15378.2.camel@localhost.localdomain> Sender: linux-bluetooth-owner@vger.kernel.org List-ID: Hi, On Mon, Sep 20, 2010, Pacho Ramos wrote: > Reporter found that the problem with his dongle was introduced in commit > aee26b30bbc24cde464ba1a557c2b258ddec6432 "Make BtIO default security > level MEDIUM", he asked here, on upstream mailing list, but didn't get > any reply clarifying this: > http://marc.info/?l=linux-bluetooth&m=127893935109510&w=2 The important piece of info missing here is an HCI trace of the failure (i.e. output of hcidump). There was a similar issue reported on IRC a few days back and the hcidump there revealed a LMP response timeout after BlueZ issues a HCI_Set_Connection_Encryption command. This would seem to indicate a bug either in the local adapter or the remote device (headset). Only an airtrace would really reveal what's going on. However, I suspect it might be possible to work around this by looking into possibilities of changing the ordering and timing of when the kernel sends the encryption request. Johan