From patchwork Tue Apr 23 17:31:39 2013 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Sebastian Andrzej Siewior X-Patchwork-Id: 238966 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 5F41C2C0115 for ; Wed, 24 Apr 2013 03:32:03 +1000 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756868Ab3DWRb6 (ORCPT ); Tue, 23 Apr 2013 13:31:58 -0400 Received: from www.linutronix.de ([62.245.132.108]:36084 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756725Ab3DWRby (ORCPT ); Tue, 23 Apr 2013 13:31:54 -0400 Received: from localhost ([127.0.0.1] helo=localhost.localdomain) by Galois.linutronix.de with esmtp (Exim 4.72) (envelope-from ) id 1UUh4L-0004gG-Lc; Tue, 23 Apr 2013 19:31:53 +0200 From: Sebastian Andrzej Siewior To: Mugunthan V N Cc: netdev@vger.kernel.org, "David S. Miller" , tglx@linutronix.de, Sebastian Andrzej Siewior Subject: [PATCH 5/5] net/cpsw: redo rx skb allocation in rx path Date: Tue, 23 Apr 2013 19:31:39 +0200 Message-Id: <1366738299-21285-6-git-send-email-bigeasy@linutronix.de> X-Mailer: git-send-email 1.7.10.4 In-Reply-To: <1366738299-21285-1-git-send-email-bigeasy@linutronix.de> References: <1366738299-21285-1-git-send-email-bigeasy@linutronix.de> X-Linutronix-Spam-Score: -1.0 X-Linutronix-Spam-Level: - X-Linutronix-Spam-Status: No , -1.0 points, 5.0 required, ALL_TRUSTED=-1, SHORTCIRCUIT=-0.0001 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org In case that we run into OOM during the allocation of the new rx-skb we don't get one and we have one skb less than we used to have. If this continues to happen then we end up with no rx-skbs at all. This patch changes the following: - if we fail to allocate the new skb, then we treat the currently completed skb as the new one and so drop the currently received data. - instead of testing multiple times if the device is gone we rely one the status field which is set to -ENOSYS in case the channel is going down and incomplete requests are purged. cpdma_chan_stop() removes most of the packages with -ENOSYS. The currently active packet which is removed has the "tear down" bit set. So if that bit is set, we send ENOSYS as well otherwise we pass the status bits which are required to figure out which of the two possible just finished. Acked-by: Mugunthan V N Signed-off-by: Sebastian Andrzej Siewior --- drivers/net/ethernet/ti/cpsw.c | 33 ++++++++++++------------------- drivers/net/ethernet/ti/davinci_cpdma.c | 7 ++++++- 2 files changed, 19 insertions(+), 21 deletions(-) diff --git a/drivers/net/ethernet/ti/cpsw.c b/drivers/net/ethernet/ti/cpsw.c index 82c4574..1fc89a5 100644 --- a/drivers/net/ethernet/ti/cpsw.c +++ b/drivers/net/ethernet/ti/cpsw.c @@ -468,43 +468,36 @@ void cpsw_tx_handler(void *token, int len, int status) void cpsw_rx_handler(void *token, int len, int status) { struct sk_buff *skb = token; + struct sk_buff *new_skb; struct net_device *ndev = skb->dev; struct cpsw_priv *priv = netdev_priv(ndev); int ret = 0; cpsw_dual_emac_src_port_detect(status, priv, ndev, skb); - /* free and bail if we are shutting down */ - if (unlikely(!netif_running(ndev)) || - unlikely(!netif_carrier_ok(ndev))) { + if (unlikely(status < 0)) { + /* the interface is going down, skbs are purged */ dev_kfree_skb_any(skb); return; } - if (likely(status >= 0)) { + + new_skb = netdev_alloc_skb_ip_align(ndev, priv->rx_packet_max); + if (new_skb) { skb_put(skb, len); cpts_rx_timestamp(priv->cpts, skb); skb->protocol = eth_type_trans(skb, ndev); netif_receive_skb(skb); priv->stats.rx_bytes += len; priv->stats.rx_packets++; - skb = NULL; - } - - if (unlikely(!netif_running(ndev))) { - if (skb) - dev_kfree_skb_any(skb); - return; + } else { + priv->stats.rx_dropped++; + new_skb = skb; } - if (likely(!skb)) { - skb = netdev_alloc_skb_ip_align(ndev, priv->rx_packet_max); - if (WARN_ON(!skb)) - return; - - ret = cpdma_chan_submit(priv->rxch, skb, skb->data, - skb_tailroom(skb), 0); - } - WARN_ON(ret < 0); + ret = cpdma_chan_submit(priv->rxch, new_skb, new_skb->data, + skb_tailroom(new_skb), 0); + if (WARN_ON(ret < 0)) + dev_kfree_skb_any(new_skb); } static irqreturn_t cpsw_interrupt(int irq, void *dev_id) diff --git a/drivers/net/ethernet/ti/davinci_cpdma.c b/drivers/net/ethernet/ti/davinci_cpdma.c index 3cc20e7..6b0a89f 100644 --- a/drivers/net/ethernet/ti/davinci_cpdma.c +++ b/drivers/net/ethernet/ti/davinci_cpdma.c @@ -776,6 +776,7 @@ static int __cpdma_chan_process(struct cpdma_chan *chan) struct cpdma_ctlr *ctlr = chan->ctlr; struct cpdma_desc __iomem *desc; int status, outlen; + int cb_status = 0; struct cpdma_desc_pool *pool = ctlr->pool; dma_addr_t desc_dma; unsigned long flags; @@ -811,8 +812,12 @@ static int __cpdma_chan_process(struct cpdma_chan *chan) } spin_unlock_irqrestore(&chan->lock, flags); + if (unlikely(status & CPDMA_DESC_TD_COMPLETE)) + cb_status = -ENOSYS; + else + cb_status = status; - __cpdma_chan_free(chan, desc, outlen, status); + __cpdma_chan_free(chan, desc, outlen, cb_status); return status; unlock_ret: