Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756050AbcJFS0s (ORCPT ); Thu, 6 Oct 2016 14:26:48 -0400 Received: from smtprelay4.synopsys.com ([198.182.47.9]:50764 "EHLO smtprelay.synopsys.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754260AbcJFS0P (ORCPT ); Thu, 6 Oct 2016 14:26:15 -0400 Subject: Re: [REGRESSION] "dwc2: gadget: fix TX FIFO size and address initialization" breaks "adb logcat" on hikey To: John Stultz , Robert Baldyga , John Youn , Felipe Balbi References: CC: lkml , Amit Pundir , Guodong Xu , "linux-usb@vger.kernel.org" From: John Youn Message-ID: Date: Thu, 6 Oct 2016 11:24:34 -0700 User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.3.0 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit X-Originating-IP: [10.9.139.95] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1297 Lines: 44 On 10/5/2016 11:33 PM, John Stultz wrote: > Hey folks, > So I've run into some trouble that I've narrowed down to commit > aa381a7259c3 ("usb: dwc2: gadget: fix TX FIFO size and address > initialization"). > > After booting up android on HiKey, if I run "adb logcat", I get about > a page of logs and then I get disconnected. On the device side, I see > the following in dmesg: > > [ 35.652076] dwc2 f72c0000.usb: dwc2_hsotg_ep_sethalt(ep > ffffffc078a4c218 ep1in, 0) > [ 35.660334] dwc2 f72c0000.usb: dwc2_hsotg_ep_sethalt(ep > ffffffc078a4c018 ep1out, 0) > > After which if I stop adbd I see: > > [ 99.471363] dwc2 f72c0000.usb: dwc2_hsotg_ep_stop_xfr: timeout flushing fifos > > and if I then restart adbd on the device it will sometimes work, but not always. > > Reverting > ba48eab8866c ("usb: dwc2: gadget: change variable name to more meaningful") > and > aa381a7259c3 ("usb: dwc2: gadget: fix TX FIFO size and address initialization") > > makes things work again. > > Any thoughts as to what might be going wrong here? I'd be happy to > help debug things. > > thanks > -john > Hi John, Yup. I've already submitted patches to revert these last week, to be applied sometime during the 4.9-rc cycle. We're working on a more robust fix for the FIFOs too. Regards, John