[meta-freescale] [meta-fsl-arm][PATCH 2/2] linux-imx-3.14.28: FEC: fix garbage send for huge data blocks

Andreas Müller schnitzeltony at googlemail.com
Mon Jun 8 05:15:31 PDT 2015


We had endless wireshark sessions here showing identical garbage contents

| (c) Copyrigt 2007-2012, Freescale Semiconductor  All rights reserved

as discussed in the thread [1] leading to the patch attached.

[1] http://www.spinics.net/lists/netdev/msg319740.html

Signed-off-by: Andreas Müller <schnitzeltony at googlemail.com>
---
 ...et-fec-fix-the-warning-found-by-dma-debug.patch | 100 +++++++++++++++++++++
 recipes-kernel/linux/linux-imx_3.14.28.bb          |   1 +
 2 files changed, 101 insertions(+)
 create mode 100644 recipes-kernel/linux/linux-imx-3.14.28/0005-Revert-net-fec-fix-the-warning-found-by-dma-debug.patch

diff --git a/recipes-kernel/linux/linux-imx-3.14.28/0005-Revert-net-fec-fix-the-warning-found-by-dma-debug.patch b/recipes-kernel/linux/linux-imx-3.14.28/0005-Revert-net-fec-fix-the-warning-found-by-dma-debug.patch
new file mode 100644
index 0000000..d40e55a
--- /dev/null
+++ b/recipes-kernel/linux/linux-imx-3.14.28/0005-Revert-net-fec-fix-the-warning-found-by-dma-debug.patch
@@ -0,0 +1,100 @@
+From a2fe37b69d4fe369c284d50927193fed81c238a0 Mon Sep 17 00:00:00 2001
+From: Fabio Estevam <fabio.estevam at freescale.com>
+Date: Fri, 13 Mar 2015 14:07:54 -0300
+Subject: [PATCH] Revert "net: fec: fix the warning found by dma debug"
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+This reverts commit 2b995f63987013bacde99168218f9c7b252bdcf1.
+
+Панов Андрей reported the following regression:
+
+"Commit 2b995f63987013bacde99168218f9c7b252bdcf1 in 4.0.0-rc3 introduces a
+nasty bug in transmit, corrupting packets.
+
+To reproduce:
+
+$ dd if=/dev/zero of=zeros bs=1M count=20
+$ md5sum -b zeros
+8f4e33f3dc3e414ff94e5fb6905cba8c *zeros
+
+This checksum is correct.
+
+Copy file "zeros" to another host with NFS, and it gets corrupted, checksum is
+changed.
+File should be big, small amounts of transmit isn't affected.
+
+I use an i.MX6 Quad board.
+
+If this commit is reverted, all works fine."
+
+Reported-by: Панов Андрей <rockford at yandex.ru>
+Signed-off-by: Fabio Estevam <fabio.estevam at freescale.com>
+Signed-off-by: David S. Miller <davem at davemloft.net>
+---
+ drivers/net/ethernet/freescale/fec_main.c | 34 ++++++++++---------------------
+ 1 file changed, 11 insertions(+), 23 deletions(-)
+
+diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c
+index 787db50..78e1ce0 100644
+--- a/drivers/net/ethernet/freescale/fec_main.c
++++ b/drivers/net/ethernet/freescale/fec_main.c
+@@ -1189,13 +1189,12 @@ static void
+ fec_enet_tx_queue(struct net_device *ndev, u16 queue_id)
+ {
+ 	struct	fec_enet_private *fep;
+-	struct bufdesc *bdp, *bdp_t;
++	struct bufdesc *bdp;
+ 	unsigned short status;
+ 	struct	sk_buff	*skb;
+ 	struct fec_enet_priv_tx_q *txq;
+ 	struct netdev_queue *nq;
+ 	int	index = 0;
+-	int	i, bdnum;
+ 	int	entries_free;
+ 
+ 	fep = netdev_priv(ndev);
+@@ -1216,29 +1215,18 @@ fec_enet_tx_queue(struct net_device *ndev, u16 queue_id)
+ 		if (bdp == txq->cur_tx)
+ 			break;
+ 
+-		bdp_t = bdp;
+-		bdnum = 1;
+-		index = fec_enet_get_bd_index(txq->tx_bd_base, bdp_t, fep);
+-		skb = txq->tx_skbuff[index];
+-		while (!skb) {
+-			bdp_t = fec_enet_get_nextdesc(bdp_t, fep, queue_id);
+-			index = fec_enet_get_bd_index(txq->tx_bd_base, bdp_t, fep);
+-			skb = txq->tx_skbuff[index];
+-			bdnum++;
+-		}
+-		if (skb_shinfo(skb)->nr_frags &&
+-		    (status = bdp_t->cbd_sc) & BD_ENET_TX_READY)
+-			break;
++		index = fec_enet_get_bd_index(txq->tx_bd_base, bdp, fep);
+ 
+-		for (i = 0; i < bdnum; i++) {
+-			if (!IS_TSO_HEADER(txq, bdp->cbd_bufaddr))
+-				dma_unmap_single(&fep->pdev->dev, bdp->cbd_bufaddr,
+-						 bdp->cbd_datlen, DMA_TO_DEVICE);
+-			bdp->cbd_bufaddr = 0;
+-			if (i < bdnum - 1)
+-				bdp = fec_enet_get_nextdesc(bdp, fep, queue_id);
+-		}
++		skb = txq->tx_skbuff[index];
+ 		txq->tx_skbuff[index] = NULL;
++		if (!IS_TSO_HEADER(txq, bdp->cbd_bufaddr))
++			dma_unmap_single(&fep->pdev->dev, bdp->cbd_bufaddr,
++					bdp->cbd_datlen, DMA_TO_DEVICE);
++		bdp->cbd_bufaddr = 0;
++		if (!skb) {
++			bdp = fec_enet_get_nextdesc(bdp, fep, queue_id);
++			continue;
++		}
+ 
+ 		/* Check for errors. */
+ 		if (status & (BD_ENET_TX_HB | BD_ENET_TX_LC |
+-- 
+1.9.3
+
diff --git a/recipes-kernel/linux/linux-imx_3.14.28.bb b/recipes-kernel/linux/linux-imx_3.14.28.bb
index 5d1be54..369c448 100644
--- a/recipes-kernel/linux/linux-imx_3.14.28.bb
+++ b/recipes-kernel/linux/linux-imx_3.14.28.bb
@@ -19,6 +19,7 @@ SRC_URI += " \
     file://0002-ARM-clk-imx6q-fix-video-divider-for-rev-T0-1.0.patch \
     file://0003-ARM-imx6sl-Disable-imx6sl-specific-code-when-imx6sl-.patch \
     file://0004-mmc-sdhci-esdhc-imx-Fixup-runtime-PM-conditions-duri.patch \
+    file://0005-ARM-imx6sl-Disable-imx6sl-specific-code-when-imx6sl-.patch \
 "
 
 COMPATIBLE_MACHINE = "(mx6)"
-- 
1.9.3



More information about the meta-freescale mailing list