From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S935224AbcIEQsc (ORCPT ); Mon, 5 Sep 2016 12:48:32 -0400 Received: from mail.linuxfoundation.org ([140.211.169.12]:52343 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S935165AbcIEQsZ (ORCPT ); Mon, 5 Sep 2016 12:48:25 -0400 From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, Brian E Rogers , Felipe Balbi Subject: [PATCH 4.7 034/143] usb: dwc3: gadget: fix for short pkts during chained xfers Date: Mon, 5 Sep 2016 18:43:30 +0200 Message-Id: <20160905164432.073159487@linuxfoundation.org> X-Mailer: git-send-email 2.9.3 In-Reply-To: <20160905164430.593075551@linuxfoundation.org> References: <20160905164430.593075551@linuxfoundation.org> User-Agent: quilt/0.64 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 4.7-stable review patch. If anyone has any objections, please let me know. ------------------ From: Felipe Balbi commit e5b36ae2f851024d43c76e51f395d32ce8d769ce upstream. DWC3 has one interesting peculiarity with chained transfers. If we setup N chained transfers and we get a short packet before processing all N TRBs, DWC3 will (conditionally) issue a XferComplete or XferInProgress event and retire all TRBs from the one which got a short packet to the last without clearing their HWO bits. This means SW must clear HWO bit manually, which this patch is doing. Cc: Brian E Rogers Signed-off-by: Felipe Balbi Signed-off-by: Greg Kroah-Hartman --- drivers/usb/dwc3/gadget.c | 23 ++++++++++++++++++++--- 1 file changed, 20 insertions(+), 3 deletions(-) --- a/drivers/usb/dwc3/gadget.c +++ b/drivers/usb/dwc3/gadget.c @@ -1897,7 +1897,8 @@ static void dwc3_gadget_free_endpoints(s static int __dwc3_cleanup_done_trbs(struct dwc3 *dwc, struct dwc3_ep *dep, struct dwc3_request *req, struct dwc3_trb *trb, - const struct dwc3_event_depevt *event, int status) + const struct dwc3_event_depevt *event, int status, + int chain) { unsigned int count; unsigned int s_pkt = 0; @@ -1905,6 +1906,19 @@ static int __dwc3_cleanup_done_trbs(stru trace_dwc3_complete_trb(dep, trb); + /* + * If we're in the middle of series of chained TRBs and we + * receive a short transfer along the way, DWC3 will skip + * through all TRBs including the last TRB in the chain (the + * where CHN bit is zero. DWC3 will also avoid clearing HWO + * bit and SW has to do it manually. + * + * We're going to do that here to avoid problems of HW trying + * to use bogus TRBs for transfers. + */ + if (chain && (trb->ctrl & DWC3_TRB_CTRL_HWO)) + trb->ctrl &= ~DWC3_TRB_CTRL_HWO; + if ((trb->ctrl & DWC3_TRB_CTRL_HWO) && status != -ESHUTDOWN) /* * We continue despite the error. There is not much we @@ -1916,6 +1930,7 @@ static int __dwc3_cleanup_done_trbs(stru */ dev_err(dwc->dev, "%s's TRB (%p) still owned by HW\n", dep->name, trb); + count = trb->size & DWC3_TRB_SIZE_MASK; if (dep->direction) { @@ -1977,10 +1992,13 @@ static int dwc3_cleanup_done_reqs(struct int ret; do { + int chain; + req = next_request(&dep->started_list); if (WARN_ON_ONCE(!req)) return 1; + chain = req->request.num_mapped_sgs > 0; i = 0; do { slot = req->first_trb_index + i; @@ -1990,9 +2008,8 @@ static int dwc3_cleanup_done_reqs(struct trb = &dep->trb_pool[slot]; count += trb->size & DWC3_TRB_SIZE_MASK; - ret = __dwc3_cleanup_done_trbs(dwc, dep, req, trb, - event, status); + event, status, chain); if (ret) break; } while (++i < req->request.num_mapped_sgs);