From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Tanya Brokhman" Subject: RE: [PATCH 5/5 v5] usb:gadget: Add SuperSpeed support to the Gadget Framework Date: Mon, 28 Mar 2011 11:15:26 +0200 Message-ID: <000601cbed28$aee4dc20$0cae9460$@org> References: <1300867498-20997-1-git-send-email-tlinder@codeaurora.org> <20110325134101.GM2609@legolas.emea.dhcp.ti.com> <000501cbed24$8efd1f70$acf75e50$@org> <20110328085412.GJ2251@legolas.emea.dhcp.ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from wolverine01.qualcomm.com ([199.106.114.254]:31889 "EHLO wolverine01.qualcomm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751231Ab1C1JOp (ORCPT ); Mon, 28 Mar 2011 05:14:45 -0400 In-Reply-To: <20110328085412.GJ2251@legolas.emea.dhcp.ti.com> Content-Language: en-us Sender: linux-arm-msm-owner@vger.kernel.org List-Id: linux-arm-msm@vger.kernel.org To: balbi@ti.com Cc: gregkh@suse.de, linux-arm-msm@vger.kernel.org, ablay@codeaurora.org, "'open list:USB GADGET/PERIPH...'" , 'open list' Hi > > I believe that in the future all gadget drivers will define their own > SS > > descriptors with values that are suitable for that gadget driver and > the > > create_ss_descriptors() will become obsolete. As I've already > mentioned this > > is just a way to allow all the existing drivers to operate in SS mode > with > > minimum changes to their code. So this patch doesn't contradict your > > approach, just extends it. > > but that's not a good approach. You're adding bloat which will be > removed soon. Problem is that once it's there, we will have to support > it for a long long time as we won't know if there are any out of tree > gadget drivers depending on that, forcing us on publising of > Documentation/feature-removal-schedule.txt that we will remove that > layer in e.g. 4 major revisions. So, IMHO, better not to add it. > I see your point. I'll generate a different patch for this function (create_ss_descriptors()) so it will be easy to take it out of the patch series if we eventually decide against it. I think it's still useful to have this ability since it can help developers at the beginning even if not adding it to the formal kernel release and just have it in the mailing list archives. Best regards, Tanya Brokhman Consultant for Qualcomm Innovation Center, Inc. Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum