Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964976AbdLVI60 (ORCPT ); Fri, 22 Dec 2017 03:58:26 -0500 Received: from mail.linuxfoundation.org ([140.211.169.12]:36184 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S964966AbdLVI6V (ORCPT ); Fri, 22 Dec 2017 03:58:21 -0500 From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, Jonas Jensen , "David S. Miller" , Sasha Levin Subject: [PATCH 4.9 063/104] net: moxa: fix TX overrun memory leak Date: Fri, 22 Dec 2017 09:46:29 +0100 Message-Id: <20171222084614.364132227@linuxfoundation.org> X-Mailer: git-send-email 2.15.1 In-Reply-To: <20171222084609.262099650@linuxfoundation.org> References: <20171222084609.262099650@linuxfoundation.org> User-Agent: quilt/0.65 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 Content-Length: 3095 Lines: 98 4.9-stable review patch. If anyone has any objections, please let me know. ------------------ From: Jonas Jensen [ Upstream commit c2b341a620018d4eaeb0e85c16274ac4e5f153d4 ] moxart_mac_start_xmit() doesn't care where tx_tail is, tx_head can catch and pass tx_tail, which is bad because moxart_tx_finished() isn't guaranteed to catch up on freeing resources from tx_tail. Add a check in moxart_mac_start_xmit() stopping the queue at the end of the circular buffer. Also add a check in moxart_tx_finished() waking the queue if the buffer has TX_WAKE_THRESHOLD or more free descriptors. While we're at it, move spin_lock_irq() to happen before our descriptor pointer is assigned in moxart_mac_start_xmit(). Addresses https://bugzilla.kernel.org/show_bug.cgi?id=99451 Signed-off-by: Jonas Jensen Signed-off-by: David S. Miller Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/moxa/moxart_ether.c | 20 ++++++++++++++++++-- drivers/net/ethernet/moxa/moxart_ether.h | 1 + 2 files changed, 19 insertions(+), 2 deletions(-) --- a/drivers/net/ethernet/moxa/moxart_ether.c +++ b/drivers/net/ethernet/moxa/moxart_ether.c @@ -25,6 +25,7 @@ #include #include #include +#include #include "moxart_ether.h" @@ -278,6 +279,13 @@ rx_next: return rx; } +static int moxart_tx_queue_space(struct net_device *ndev) +{ + struct moxart_mac_priv_t *priv = netdev_priv(ndev); + + return CIRC_SPACE(priv->tx_head, priv->tx_tail, TX_DESC_NUM); +} + static void moxart_tx_finished(struct net_device *ndev) { struct moxart_mac_priv_t *priv = netdev_priv(ndev); @@ -297,6 +305,9 @@ static void moxart_tx_finished(struct ne tx_tail = TX_NEXT(tx_tail); } priv->tx_tail = tx_tail; + if (netif_queue_stopped(ndev) && + moxart_tx_queue_space(ndev) >= TX_WAKE_THRESHOLD) + netif_wake_queue(ndev); } static irqreturn_t moxart_mac_interrupt(int irq, void *dev_id) @@ -324,13 +335,18 @@ static int moxart_mac_start_xmit(struct struct moxart_mac_priv_t *priv = netdev_priv(ndev); void *desc; unsigned int len; - unsigned int tx_head = priv->tx_head; + unsigned int tx_head; u32 txdes1; int ret = NETDEV_TX_BUSY; + spin_lock_irq(&priv->txlock); + + tx_head = priv->tx_head; desc = priv->tx_desc_base + (TX_REG_DESC_SIZE * tx_head); - spin_lock_irq(&priv->txlock); + if (moxart_tx_queue_space(ndev) == 1) + netif_stop_queue(ndev); + if (moxart_desc_read(desc + TX_REG_OFFSET_DESC0) & TX_DESC0_DMA_OWN) { net_dbg_ratelimited("no TX space for packet\n"); priv->stats.tx_dropped++; --- a/drivers/net/ethernet/moxa/moxart_ether.h +++ b/drivers/net/ethernet/moxa/moxart_ether.h @@ -59,6 +59,7 @@ #define TX_NEXT(N) (((N) + 1) & (TX_DESC_NUM_MASK)) #define TX_BUF_SIZE 1600 #define TX_BUF_SIZE_MAX (TX_DESC1_BUF_SIZE_MASK+1) +#define TX_WAKE_THRESHOLD 16 #define RX_DESC_NUM 64 #define RX_DESC_NUM_MASK (RX_DESC_NUM-1)