Netdev List
 help / color / mirror / Atom feed
From: Claudiu Manoil <claudiu.manoil@freescale.com>
To: <netdev@vger.kernel.org>
Cc: "David S. Miller" <davem@davemloft.net>
Subject: [PATCH net-next v2 6/6] gianfar: Remove clean_rx_ring race from gfar_ethtool
Date: Mon, 17 Feb 2014 12:53:19 +0200	[thread overview]
Message-ID: <1392634399-5456-7-git-send-email-claudiu.manoil@freescale.com> (raw)
In-Reply-To: <1392634399-5456-1-git-send-email-claudiu.manoil@freescale.com>

gfar_clean_rx_ring() was designed to be called from napi
(rx softirq) context to do the Rx processing. Calling it
from a process context like this is a bug as it will
clearly race with the napi Rx processing.

There's also no point in initializing num_txbdfree since
startup_gfar() already does that, when bringing the device
up again (after reset). Changing num_txbdfree "on-the-fly"
like this is also subject to race conditions.  num_txbdfree
is handled by the Tx processing path and the device reset
procedure.  Also, don't assume that num_rx_queues is always
equal to num_tx_queues.

Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
---
 drivers/net/ethernet/freescale/gianfar_ethtool.c | 63 +++---------------------
 1 file changed, 8 insertions(+), 55 deletions(-)

diff --git a/drivers/net/ethernet/freescale/gianfar_ethtool.c b/drivers/net/ethernet/freescale/gianfar_ethtool.c
index 69fab72..19557ec 100644
--- a/drivers/net/ethernet/freescale/gianfar_ethtool.c
+++ b/drivers/net/ethernet/freescale/gianfar_ethtool.c
@@ -44,9 +44,6 @@
 
 #include "gianfar.h"
 
-extern int gfar_clean_rx_ring(struct gfar_priv_rx_q *rx_queue,
-			      int rx_work_limit);
-
 #define GFAR_MAX_COAL_USECS 0xffff
 #define GFAR_MAX_COAL_FRAMES 0xff
 static void gfar_fill_stats(struct net_device *dev, struct ethtool_stats *dummy,
@@ -466,15 +463,13 @@ static void gfar_gringparam(struct net_device *dev,
 }
 
 /* Change the current ring parameters, stopping the controller if
- * necessary so that we don't mess things up while we're in
- * motion.  We wait for the ring to be clean before reallocating
- * the rings.
+ * necessary so that we don't mess things up while we're in motion.
  */
 static int gfar_sringparam(struct net_device *dev,
 			   struct ethtool_ringparam *rvals)
 {
 	struct gfar_private *priv = netdev_priv(dev);
-	int err = 0, i = 0;
+	int err = 0, i;
 
 	if (rvals->rx_pending > GFAR_RX_MAX_RING_SIZE)
 		return -EINVAL;
@@ -492,38 +487,15 @@ static int gfar_sringparam(struct net_device *dev,
 		return -EINVAL;
 	}
 
-
-	if (dev->flags & IFF_UP) {
-		unsigned long flags;
-
-		/* Halt TX and RX, and process the frames which
-		 * have already been received
-		 */
-		local_irq_save(flags);
-		lock_tx_qs(priv);
-		lock_rx_qs(priv);
-
-		gfar_halt(priv);
-
-		unlock_rx_qs(priv);
-		unlock_tx_qs(priv);
-		local_irq_restore(flags);
-
-		for (i = 0; i < priv->num_rx_queues; i++)
-			gfar_clean_rx_ring(priv->rx_queue[i],
-					   priv->rx_queue[i]->rx_ring_size);
-
-		/* Now we take down the rings to rebuild them */
+	if (dev->flags & IFF_UP)
 		stop_gfar(dev);
-	}
 
-	/* Change the size */
-	for (i = 0; i < priv->num_rx_queues; i++) {
+	/* Change the sizes */
+	for (i = 0; i < priv->num_rx_queues; i++)
 		priv->rx_queue[i]->rx_ring_size = rvals->rx_pending;
+
+	for (i = 0; i < priv->num_tx_queues; i++)
 		priv->tx_queue[i]->tx_ring_size = rvals->tx_pending;
-		priv->tx_queue[i]->num_txbdfree =
-			priv->tx_queue[i]->tx_ring_size;
-	}
 
 	/* Rebuild the rings with the new size */
 	if (dev->flags & IFF_UP) {
@@ -607,10 +579,8 @@ static int gfar_spauseparam(struct net_device *dev,
 
 int gfar_set_features(struct net_device *dev, netdev_features_t features)
 {
-	struct gfar_private *priv = netdev_priv(dev);
-	unsigned long flags;
-	int err = 0, i = 0;
 	netdev_features_t changed = dev->features ^ features;
+	int err = 0;
 
 	if (changed & (NETIF_F_HW_VLAN_CTAG_TX|NETIF_F_HW_VLAN_CTAG_RX))
 		gfar_vlan_mode(dev, features);
@@ -619,23 +589,6 @@ int gfar_set_features(struct net_device *dev, netdev_features_t features)
 		return 0;
 
 	if (dev->flags & IFF_UP) {
-		/* Halt TX and RX, and process the frames which
-		 * have already been received
-		 */
-		local_irq_save(flags);
-		lock_tx_qs(priv);
-		lock_rx_qs(priv);
-
-		gfar_halt(priv);
-
-		unlock_tx_qs(priv);
-		unlock_rx_qs(priv);
-		local_irq_restore(flags);
-
-		for (i = 0; i < priv->num_rx_queues; i++)
-			gfar_clean_rx_ring(priv->rx_queue[i],
-					   priv->rx_queue[i]->rx_ring_size);
-
 		/* Now we take down the rings to rebuild them */
 		stop_gfar(dev);
 
-- 
1.7.11.7

  parent reply	other threads:[~2014-02-17 10:53 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-17 10:53 [PATCH net-next v2 0/6] gianfar: Device configuration fixes Claudiu Manoil
2014-02-17 10:53 ` [PATCH net-next v2 1/6] gianfar: Cleanup/Fix gfar_probe and the hw init code Claudiu Manoil
2014-02-17 10:53 ` [PATCH net-next v2 2/6] gianfar: Remove sysfs stubs for FIFOCFG and stashing Claudiu Manoil
2014-02-17 10:53 ` [PATCH net-next v2 3/6] gianfar: Remove useless HAS_PADDING device flag Claudiu Manoil
2014-02-17 10:53 ` [PATCH net-next v2 4/6] gianfar: Factor out enabling/disabling of hw interrupts Claudiu Manoil
2014-02-17 10:53 ` [PATCH net-next v2 5/6] gianfar: Add missing graceful reset steps and fixes Claudiu Manoil
2014-02-17 10:53 ` Claudiu Manoil [this message]
2014-02-18 20:03 ` [PATCH net-next v2 0/6] gianfar: Device configuration fixes David Miller

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1392634399-5456-7-git-send-email-claudiu.manoil@freescale.com \
    --to=claudiu.manoil@freescale.com \
    --cc=davem@davemloft.net \
    --cc=netdev@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox