Return-Path: Message-ID: <4DD5F955.20305@Atheros.com> Date: Fri, 20 May 2011 10:47:09 +0530 From: Suraj Sumangala MIME-Version: 1.0 To: Mat Martineau , Arun Kumar SINGH , Anurag Gupta , Peter Krystad , Subject: Re: BT 3.0 HS Support in BlueZ References: <20110419180425.GA2476@joana> <20110505201531.GE2098@joana> <20110509225246.GB2203@joana> <4DD4AB14.9050401@Atheros.com> <20110519184018.GA4888@joana> In-Reply-To: <20110519184018.GA4888@joana> Content-Type: text/plain; charset="ISO-8859-1"; format=flowed Sender: linux-bluetooth-owner@vger.kernel.org List-ID: Hi Mat, On 5/20/2011 12:10 AM, Gustavo F. Padovan wrote: > Hi Mat, > > * Mat Martineau [2011-05-19 09:58:04 -0700]: > >> >> >> On Thu, 19 May 2011, Suraj Sumangala wrote: >> >>> Hi Mat, >>> >>> On 5/19/2011 2:28 AM, Mat Martineau wrote: >>>> >>>> Gustavo, >>>> >>>> On Mon, 9 May 2011, Gustavo F. Padovan wrote: >>>> >> >> ... snip ... >> >>>>> >>>>> Isn't Extended Flow Specification a required feature for AMP? >>>>> I haven't seen >>>>> it in your implementation. >>>> >>>> Extended Flowspec is needed to create an L2CAP channel directly on >>>> AMP, but the implementation you're looking at does not implement the >>>> "create channel" feature. Channels are created on BR/EDR and moved >>>> to AMP, which does not require extended flowspec. >>>> >>> >>> Why don't we have to use EFS for channels moved from BDR? Is it >>> because we assume that the QoS provided by AMP will be better than >>> BDR? >> >> Only "Best Effort" is supported by this implementation so QoS is >> equivalent on either controller type. EFS is optional when creating >> channels on BR/EDR, and the spec does not require EFS when moving to >> a Best Effort AMP link. We've extensively interop'd AMP channel >> moves without EFS.g > > Isn't EFS and Create Channel Command required to the qualification? > In PTS when AMP Support and AMP Manager Channel are enabled tests > for Create Channel and EFS are also enabled. > I also see test cases which require you to do L2CAP configure with EFS and 'Best Effort' as the service type, Aren't they mandatory test cases? But, I think what you said is correct. Here is what the spec say about it. "Since the Identifier for an Extended Flow Specification with Service Type Best Effort is fixed to 0x01 it is possible to generate a Best Effort Extended Flow Specification for the remote device without performing the Lockstep Configuration process" Regards Suraj