[PATCH 0/3] ucc_geth: Some more fixes

[PATCH 0/3] ucc_geth: Some more fixes

[Anton Vorontsov]

Hi all,

Here are some more fixes and one small enhancement for the ucc_geth
driver.

Thanks,

-- 
Anton Vorontsov
email: cbouatmailru@gmail.com
irc://irc.freenode.net/bd2

[PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Anton Vorontsov]

Following oops was seen with the ucc_geth driver:

 Unable to handle kernel paging request for data at address 0x00000058
 Faulting instruction address: 0xc024f2fc
 Oops: Kernel access of bad area, sig: 11 [#1]
 [...]
 NIP [c024f2fc] skb_recycle_check+0x14/0x100
 LR [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
 Call Trace:
 [df857d50] [c000b03c] __ipipe_grab_irq+0x3c/0xa4 (unreliable)
 [df857d60] [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
 [df857dd0] [c0258cf8] net_rx_action+0xf8/0x1b8
 [df857e10] [c0032a38] __do_softirq+0xb8/0x13c
 [df857e60] [c00065cc] do_softirq+0xa0/0xac
 [...]

This is because ucc_geth_tx() tries to process an empty queue when
queues are logically stopped. Stopping the queues doesn't disable
polling, and since nowadays ucc_geth_tx() is actually called from
the polling routine, the oops above might pop up.

Fix this by removing 'netif_queue_stopped() == 0' check.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
---
 drivers/net/ucc_geth.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index afaf088..0f8c99e 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3273,7 +3273,7 @@ static int ucc_geth_tx(struct net_device *dev, u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if ((bd == ugeth->txBd[txQ]) && (netif_queue_stopped(dev) == 0))
+		if (bd == ugeth->txBd[txQ]) /* queue empty? */
 			break;
 
 		dev->stats.tx_packets++;
-- 
1.6.3.3

[PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Anton Vorontsov]

Following oops was seen with the ucc_geth driver:

 Unable to handle kernel paging request for data at address 0x00000058
 Faulting instruction address: 0xc024f2fc
 Oops: Kernel access of bad area, sig: 11 [#1]
 [...]
 NIP [c024f2fc] skb_recycle_check+0x14/0x100
 LR [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
 Call Trace:
 [df857d50] [c000b03c] __ipipe_grab_irq+0x3c/0xa4 (unreliable)
 [df857d60] [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
 [df857dd0] [c0258cf8] net_rx_action+0xf8/0x1b8
 [df857e10] [c0032a38] __do_softirq+0xb8/0x13c
 [df857e60] [c00065cc] do_softirq+0xa0/0xac
 [...]

This is because ucc_geth_tx() tries to process an empty queue when
queues are logically stopped. Stopping the queues doesn't disable
polling, and since nowadays ucc_geth_tx() is actually called from
the polling routine, the oops above might pop up.

Fix this by removing 'netif_queue_stopped() == 0' check.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
---
 drivers/net/ucc_geth.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index afaf088..0f8c99e 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3273,7 +3273,7 @@ static int ucc_geth_tx(struct net_device *dev, u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if ((bd == ugeth->txBd[txQ]) && (netif_queue_stopped(dev) == 0))
+		if (bd == ugeth->txBd[txQ]) /* queue empty? */
 			break;
 
 		dev->stats.tx_packets++;
-- 
1.6.3.3

[PATCH 2/3] ucc_geth: Fix netdev watchdog triggering on link changes

Since commit 864fdf884e82bacbe8ca5e93bd43393a61d2e2b4 ("ucc_geth:
Fix hangs after switching from full to half duplex") ucc_geth driver
disables the controller during MAC configuration changes. Though,
disabling the controller might take quite awhile, and so the netdev
watchdog might get upset:

 NETDEV WATCHDOG: eth2 (ucc_geth): transmit queue 0 timed out
 ------------[ cut here ]------------
 Badness at c02729a8 [verbose debug info unavailable]
 NIP: c02729a8 LR: c02729a8 CTR: c01b6088
 REGS: c0451c40 TRAP: 0700   Not tainted  (2.6.32-trunk-8360e)
 [...]
 NIP [c02729a8] dev_watchdog+0x280/0x290
 LR [c02729a8] dev_watchdog+0x280/0x290
 Call Trace:
 [c0451cf0] [c02729a8] dev_watchdog+0x280/0x290 (unreliable)
 [c0451d50] [c00377c4] run_timer_softirq+0x164/0x224
 [c0451da0] [c0032a38] __do_softirq+0xb8/0x13c
 [c0451df0] [c00065cc] do_softirq+0xa0/0xac
 [c0451e00] [c003280c] irq_exit+0x7c/0x9c
 [c0451e10] [c00640c4] __ipipe_sync_stage+0x248/0x24c
 [...]

This patch fixes the issue by detaching the netdev during the
time we change the configuration.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
---
 drivers/net/ucc_geth.c |    7 +++++--
 1 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 0f8c99e..7fff4c5 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -1563,7 +1563,10 @@ static int ugeth_disable(struct ucc_geth_private *ugeth, enum comm_dir mode)
 
 static void ugeth_quiesce(struct ucc_geth_private *ugeth)
 {
-	/* Wait for and prevent any further xmits. */
+	/* Prevent any further xmits, plus detach the device. */
+	netif_device_detach(ugeth->ndev);
+
+	/* Wait for any current xmits to finish. */
 	netif_tx_disable(ugeth->ndev);
 
 	/* Disable the interrupt to avoid NAPI rescheduling. */
@@ -1577,7 +1580,7 @@ static void ugeth_activate(struct ucc_geth_private *ugeth)
 {
 	napi_enable(&ugeth->napi);
 	enable_irq(ugeth->ug_info->uf_info.irq);
-	netif_tx_wake_all_queues(ugeth->ndev);
+	netif_device_attach(ugeth->ndev);
 }
 
 /* Called every time the controller might need to be made
-- 
1.6.3.3

[PATCH 2/3] ucc_geth: Fix netdev watchdog triggering on link changes

[Anton Vorontsov]

Since commit 864fdf884e82bacbe8ca5e93bd43393a61d2e2b4 ("ucc_geth:
Fix hangs after switching from full to half duplex") ucc_geth driver
disables the controller during MAC configuration changes. Though,
disabling the controller might take quite awhile, and so the netdev
watchdog might get upset:

 NETDEV WATCHDOG: eth2 (ucc_geth): transmit queue 0 timed out
 ------------[ cut here ]------------
 Badness at c02729a8 [verbose debug info unavailable]
 NIP: c02729a8 LR: c02729a8 CTR: c01b6088
 REGS: c0451c40 TRAP: 0700   Not tainted  (2.6.32-trunk-8360e)
 [...]
 NIP [c02729a8] dev_watchdog+0x280/0x290
 LR [c02729a8] dev_watchdog+0x280/0x290
 Call Trace:
 [c0451cf0] [c02729a8] dev_watchdog+0x280/0x290 (unreliable)
 [c0451d50] [c00377c4] run_timer_softirq+0x164/0x224
 [c0451da0] [c0032a38] __do_softirq+0xb8/0x13c
 [c0451df0] [c00065cc] do_softirq+0xa0/0xac
 [c0451e00] [c003280c] irq_exit+0x7c/0x9c
 [c0451e10] [c00640c4] __ipipe_sync_stage+0x248/0x24c
 [...]

This patch fixes the issue by detaching the netdev during the
time we change the configuration.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
---
 drivers/net/ucc_geth.c |    7 +++++--
 1 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 0f8c99e..7fff4c5 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -1563,7 +1563,10 @@ static int ugeth_disable(struct ucc_geth_private *ugeth, enum comm_dir mode)
 
 static void ugeth_quiesce(struct ucc_geth_private *ugeth)
 {
-	/* Wait for and prevent any further xmits. */
+	/* Prevent any further xmits, plus detach the device. */
+	netif_device_detach(ugeth->ndev);
+
+	/* Wait for any current xmits to finish. */
 	netif_tx_disable(ugeth->ndev);
 
 	/* Disable the interrupt to avoid NAPI rescheduling. */
@@ -1577,7 +1580,7 @@ static void ugeth_activate(struct ucc_geth_private *ugeth)
 {
 	napi_enable(&ugeth->napi);
 	enable_irq(ugeth->ug_info->uf_info.irq);
-	netif_tx_wake_all_queues(ugeth->ndev);
+	netif_device_attach(ugeth->ndev);
 }
 
 /* Called every time the controller might need to be made
-- 
1.6.3.3

[PATCH 3/3] ucc_geth: Don't needlessly change MAC settings in adjust_link()

[Anton Vorontsov]

If PHY doesn't have an IRQ, phylib would poll for link changes, and
would call adjust_link() every second. In that case we disable and
enable the controller every second.

Let's better check if there is actually anything changed, and, if so,
change the MAC settings.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
---
 drivers/net/ucc_geth.c |   33 ++++++++++++++++++---------------
 1 files changed, 18 insertions(+), 15 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 7fff4c5..41ad2f3 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -1651,25 +1651,28 @@ static void adjust_link(struct net_device *dev)
 			ugeth->oldspeed = phydev->speed;
 		}
 
-		/*
-		 * To change the MAC configuration we need to disable the
-		 * controller. To do so, we have to either grab ugeth->lock,
-		 * which is a bad idea since 'graceful stop' commands might
-		 * take quite a while, or we can quiesce driver's activity.
-		 */
-		ugeth_quiesce(ugeth);
-		ugeth_disable(ugeth, COMM_DIR_RX_AND_TX);
-
-		out_be32(&ug_regs->maccfg2, tempval);
-		out_be32(&uf_regs->upsmr, upsmr);
-
-		ugeth_enable(ugeth, COMM_DIR_RX_AND_TX);
-		ugeth_activate(ugeth);
-
 		if (!ugeth->oldlink) {
 			new_state = 1;
 			ugeth->oldlink = 1;
 		}
+
+		if (new_state) {
+			/*
+			 * To change the MAC configuration we need to disable
+			 * the controller. To do so, we have to either grab
+			 * ugeth->lock, which is a bad idea since 'graceful
+			 * stop' commands might take quite a while, or we can
+			 * quiesce driver's activity.
+			 */
+			ugeth_quiesce(ugeth);
+			ugeth_disable(ugeth, COMM_DIR_RX_AND_TX);
+
+			out_be32(&ug_regs->maccfg2, tempval);
+			out_be32(&uf_regs->upsmr, upsmr);
+
+			ugeth_enable(ugeth, COMM_DIR_RX_AND_TX);
+			ugeth_activate(ugeth);
+		}
 	} else if (ugeth->oldlink) {
 			new_state = 1;
 			ugeth->oldlink = 0;
-- 
1.6.3.3

[PATCH 3/3] ucc_geth: Don't needlessly change MAC settings in adjust_link()

[Anton Vorontsov]

If PHY doesn't have an IRQ, phylib would poll for link changes, and
would call adjust_link() every second. In that case we disable and
enable the controller every second.

Let's better check if there is actually anything changed, and, if so,
change the MAC settings.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
---
 drivers/net/ucc_geth.c |   33 ++++++++++++++++++---------------
 1 files changed, 18 insertions(+), 15 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 7fff4c5..41ad2f3 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -1651,25 +1651,28 @@ static void adjust_link(struct net_device *dev)
 			ugeth->oldspeed = phydev->speed;
 		}
 
-		/*
-		 * To change the MAC configuration we need to disable the
-		 * controller. To do so, we have to either grab ugeth->lock,
-		 * which is a bad idea since 'graceful stop' commands might
-		 * take quite a while, or we can quiesce driver's activity.
-		 */
-		ugeth_quiesce(ugeth);
-		ugeth_disable(ugeth, COMM_DIR_RX_AND_TX);
-
-		out_be32(&ug_regs->maccfg2, tempval);
-		out_be32(&uf_regs->upsmr, upsmr);
-
-		ugeth_enable(ugeth, COMM_DIR_RX_AND_TX);
-		ugeth_activate(ugeth);
-
 		if (!ugeth->oldlink) {
 			new_state = 1;
 			ugeth->oldlink = 1;
 		}
+
+		if (new_state) {
+			/*
+			 * To change the MAC configuration we need to disable
+			 * the controller. To do so, we have to either grab
+			 * ugeth->lock, which is a bad idea since 'graceful
+			 * stop' commands might take quite a while, or we can
+			 * quiesce driver's activity.
+			 */
+			ugeth_quiesce(ugeth);
+			ugeth_disable(ugeth, COMM_DIR_RX_AND_TX);
+
+			out_be32(&ug_regs->maccfg2, tempval);
+			out_be32(&uf_regs->upsmr, upsmr);
+
+			ugeth_enable(ugeth, COMM_DIR_RX_AND_TX);
+			ugeth_activate(ugeth);
+		}
 	} else if (ugeth->oldlink) {
 			new_state = 1;
 			ugeth->oldlink = 0;
-- 
1.6.3.3

Re: [PATCH 0/3] ucc_geth: Some more fixes

[David Miller]

From: Anton Vorontsov <avorontsov@ru.mvista.com>
Date: Thu, 24 Dec 2009 18:30:34 +0300

> Here are some more fixes and one small enhancement for the ucc_geth
> driver.

All 3 patches applied, thanks a lot!

Re: [PATCH 0/3] ucc_geth: Some more fixes

[David Miller]

From: Anton Vorontsov <avorontsov@ru.mvista.com>
Date: Thu, 24 Dec 2009 18:30:34 +0300

> Here are some more fixes and one small enhancement for the ucc_geth
> driver.

All 3 patches applied, thanks a lot!

RE: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Wu Jiajun-B06378]

=20
'bd =3D=3D ugeth->txBd[txQ]' has two possible statuses: 1)full queue.
2)empty queue.
Removing 'netif_queue_stopped() =3D=3D 0' will make transmitting =
stopping
when the queue is full.=20

I made a new patch for this oops.

>From 726765194352d01dc8ea672d97612589b67cec3f Mon Sep 17 00:00:00 2001
From: Jiajun Wu <b06378@freescale.com>
Date: Mon, 11 Jan 2010 10:57:22 +0800
Subject: [PATCH] ucc_geth: Fix empty TX queue processing

Signed-off-by: Jiajun Wu <b06378@freescale.com>
---
 drivers/net/ucc_geth.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index f982220..34345f0 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3576,17 +3576,17 @@ static int ucc_geth_tx(struct net_device *dev,
u8 txQ)
 	while ((bd_status & T_R) =3D=3D 0) {
 		struct sk_buff *skb;
=20
+		skb =3D ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
+
 		/* BD contains already transmitted buffer.   */
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
=20
-		if ((bd =3D=3D ugeth->txBd[txQ]) &&
(netif_queue_stopped(dev) =3D=3D 0))
+		if ((bd =3D=3D ugeth->txBd[txQ]) && (skb =3D=3D NULL))
 			break;
=20
 		dev->stats.tx_packets++;
=20
-		skb =3D ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
-
 		if (skb_queue_len(&ugeth->rx_recycle) < RX_BD_RING_LEN
&&
 			     skb_recycle_check(skb,
=20
ugeth->ug_info->uf_info.max_rx_buf_length +
--=20
1.5.6.3
-----Original Message-----
From: linuxppc-dev-bounces+b13201=3Dfreescale.com@lists.ozlabs.org
[mailto:linuxppc-dev-bounces+b13201=3Dfreescale.com@lists.ozlabs.org] On
Behalf Of Anton Vorontsov
Sent: Thursday, December 24, 2009 11:31 PM
To: David Miller
Cc: linuxppc-dev@ozlabs.org; netdev@vger.kernel.org; Lennart Sorensen
Subject: [PATCH 1/3] ucc_geth: Fix empty TX queue processing
-----------

Following oops was seen with the ucc_geth driver:

 Unable to handle kernel paging request for data at address 0x00000058
Faulting instruction address: 0xc024f2fc
 Oops: Kernel access of bad area, sig: 11 [#1]  [...]  NIP [c024f2fc]
skb_recycle_check+0x14/0x100  LR [e30aa0a4] ucc_geth_poll+0xd8/0x4e0
[ucc_geth_driver]  Call Trace:
 [df857d50] [c000b03c] __ipipe_grab_irq+0x3c/0xa4 (unreliable)
[df857d60] [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
[df857dd0] [c0258cf8] net_rx_action+0xf8/0x1b8  [df857e10] [c0032a38]
__do_softirq+0xb8/0x13c  [df857e60] [c00065cc] do_softirq+0xa0/0xac
[...]

This is because ucc_geth_tx() tries to process an empty queue when
queues are logically stopped. Stopping the queues doesn't disable
polling, and since nowadays ucc_geth_tx() is actually called from the
polling routine, the oops above might pop up.

Fix this by removing 'netif_queue_stopped() =3D=3D 0' check.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
---
 drivers/net/ucc_geth.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c index
afaf088..0f8c99e 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3273,7 +3273,7 @@ static int ucc_geth_tx(struct net_device *dev, u8
txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
=20
-		if ((bd =3D=3D ugeth->txBd[txQ]) &&
(netif_queue_stopped(dev) =3D=3D 0))
+		if (bd =3D=3D ugeth->txBd[txQ]) /* queue empty? */
 			break;
=20
 		dev->stats.tx_packets++;
--
1.6.3.3

_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

RE: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Wu Jiajun-B06378]

 
'bd == ugeth->txBd[txQ]' has two possible statuses: 1)full queue.
2)empty queue.
Removing 'netif_queue_stopped() == 0' will make transmitting stopping
when the queue is full. 

I made a new patch for this oops.

>From 726765194352d01dc8ea672d97612589b67cec3f Mon Sep 17 00:00:00 2001
From: Jiajun Wu <b06378@freescale.com>
Date: Mon, 11 Jan 2010 10:57:22 +0800
Subject: [PATCH] ucc_geth: Fix empty TX queue processing

Signed-off-by: Jiajun Wu <b06378@freescale.com>
---
 drivers/net/ucc_geth.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index f982220..34345f0 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3576,17 +3576,17 @@ static int ucc_geth_tx(struct net_device *dev,
u8 txQ)
 	while ((bd_status & T_R) == 0) {
 		struct sk_buff *skb;
 
+		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
+
 		/* BD contains already transmitted buffer.   */
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if ((bd == ugeth->txBd[txQ]) &&
(netif_queue_stopped(dev) == 0))
+		if ((bd == ugeth->txBd[txQ]) && (skb == NULL))
 			break;
 
 		dev->stats.tx_packets++;
 
-		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
-
 		if (skb_queue_len(&ugeth->rx_recycle) < RX_BD_RING_LEN
&&
 			     skb_recycle_check(skb,
 
ugeth->ug_info->uf_info.max_rx_buf_length +
-- 
1.5.6.3
-----Original Message-----
From: linuxppc-dev-bounces+b13201=freescale.com@lists.ozlabs.org
[mailto:linuxppc-dev-bounces+b13201=freescale.com@lists.ozlabs.org] On
Behalf Of Anton Vorontsov
Sent: Thursday, December 24, 2009 11:31 PM
To: David Miller
Cc: linuxppc-dev@ozlabs.org; netdev@vger.kernel.org; Lennart Sorensen
Subject: [PATCH 1/3] ucc_geth: Fix empty TX queue processing
-----------

Following oops was seen with the ucc_geth driver:

 Unable to handle kernel paging request for data at address 0x00000058
Faulting instruction address: 0xc024f2fc
 Oops: Kernel access of bad area, sig: 11 [#1]  [...]  NIP [c024f2fc]
skb_recycle_check+0x14/0x100  LR [e30aa0a4] ucc_geth_poll+0xd8/0x4e0
[ucc_geth_driver]  Call Trace:
 [df857d50] [c000b03c] __ipipe_grab_irq+0x3c/0xa4 (unreliable)
[df857d60] [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
[df857dd0] [c0258cf8] net_rx_action+0xf8/0x1b8  [df857e10] [c0032a38]
__do_softirq+0xb8/0x13c  [df857e60] [c00065cc] do_softirq+0xa0/0xac
[...]

This is because ucc_geth_tx() tries to process an empty queue when
queues are logically stopped. Stopping the queues doesn't disable
polling, and since nowadays ucc_geth_tx() is actually called from the
polling routine, the oops above might pop up.

Fix this by removing 'netif_queue_stopped() == 0' check.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
---
 drivers/net/ucc_geth.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c index
afaf088..0f8c99e 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3273,7 +3273,7 @@ static int ucc_geth_tx(struct net_device *dev, u8
txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if ((bd == ugeth->txBd[txQ]) &&
(netif_queue_stopped(dev) == 0))
+		if (bd == ugeth->txBd[txQ]) /* queue empty? */
 			break;
 
 		dev->stats.tx_packets++;
--
1.6.3.3

_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Re: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Anton Vorontsov]

On Mon, Jan 11, 2010 at 11:47:37AM +0800, Wu Jiajun-B06378 wrote:
>  
> 'bd == ugeth->txBd[txQ]' has two possible statuses: 1)full queue.
> 2)empty queue.
> Removing 'netif_queue_stopped() == 0' will make transmitting stopping
> when the queue is full. 
> 
> I made a new patch for this oops.
[...]
> +		if ((bd == ugeth->txBd[txQ]) && (skb == NULL))
>  			break;

Hm. I wonder why do we need the 'bd == ugeth->txBd[txQ]' check
at all? The null skb will cause a kernel oops anyway.

I think the patch below should be sufficient for the fix.
Can you try it? Or if you tell me how to reproduce the issue
you observe, I can try it myself.

Thanks a lot!

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 41ad2f3..a1a6d06 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3279,13 +3279,12 @@ static int ucc_geth_tx(struct net_device *dev, u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if (bd == ugeth->txBd[txQ]) /* queue empty? */
+		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
+		if (!skb)
 			break;
 
 		dev->stats.tx_packets++;
 
-		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
-
 		if (skb_queue_len(&ugeth->rx_recycle) < RX_BD_RING_LEN &&
 			     skb_recycle_check(skb,
 				    ugeth->ug_info->uf_info.max_rx_buf_length +

Re: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Anton Vorontsov]

On Mon, Jan 11, 2010 at 11:47:37AM +0800, Wu Jiajun-B06378 wrote:
>  
> 'bd == ugeth->txBd[txQ]' has two possible statuses: 1)full queue.
> 2)empty queue.
> Removing 'netif_queue_stopped() == 0' will make transmitting stopping
> when the queue is full. 
> 
> I made a new patch for this oops.
[...]
> +		if ((bd == ugeth->txBd[txQ]) && (skb == NULL))
>  			break;

Hm. I wonder why do we need the 'bd == ugeth->txBd[txQ]' check
at all? The null skb will cause a kernel oops anyway.

I think the patch below should be sufficient for the fix.
Can you try it? Or if you tell me how to reproduce the issue
you observe, I can try it myself.

Thanks a lot!

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 41ad2f3..a1a6d06 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3279,13 +3279,12 @@ static int ucc_geth_tx(struct net_device *dev, u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if (bd == ugeth->txBd[txQ]) /* queue empty? */
+		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
+		if (!skb)
 			break;
 
 		dev->stats.tx_packets++;
 
-		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
-
 		if (skb_queue_len(&ugeth->rx_recycle) < RX_BD_RING_LEN &&
 			     skb_recycle_check(skb,
 				    ugeth->ug_info->uf_info.max_rx_buf_length +

RE: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Wu Jiajun-B06378]

Yes,'if (!skb)' is enough.
You can reproduce transmitting stopping if you use 'if ((bd =3D=3D =
ugeth->txBd[txQ])' and run ipforwarding with MTU=3D64 1Gbps =
100%linerate.


-----Original Message-----
From: Anton Vorontsov [mailto:avorontsov@ru.mvista.com]=20
Sent: 2010=C4=EA1=D4=C211=C8=D5 18:53
To: Wu Jiajun-B06378
Cc: linuxppc-dev@ozlabs.org; netdev@vger.kernel.org; =
lsorense@csclub.uwaterloo.ca; davem@davemloft.net
Subject: Re: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

On Mon, Jan 11, 2010 at 11:47:37AM +0800, Wu Jiajun-B06378 wrote:
> =20
> 'bd =3D=3D ugeth->txBd[txQ]' has two possible statuses: 1)full queue.
> 2)empty queue.
> Removing 'netif_queue_stopped() =3D=3D 0' will make transmitting =
stopping=20
> when the queue is full.
>=20
> I made a new patch for this oops.
[...]
> +		if ((bd =3D=3D ugeth->txBd[txQ]) && (skb =3D=3D NULL))
>  			break;

Hm. I wonder why do we need the 'bd =3D=3D ugeth->txBd[txQ]' check at =
all? The null skb will cause a kernel oops anyway.

I think the patch below should be sufficient for the fix.
Can you try it? Or if you tell me how to reproduce the issue you =
observe, I can try it myself.

Thanks a lot!

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c index =
41ad2f3..a1a6d06 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3279,13 +3279,12 @@ static int ucc_geth_tx(struct net_device *dev, =
u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
=20
-		if (bd =3D=3D ugeth->txBd[txQ]) /* queue empty? */
+		skb =3D ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
+		if (!skb)
 			break;
=20
 		dev->stats.tx_packets++;
=20
-		skb =3D ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
-
 		if (skb_queue_len(&ugeth->rx_recycle) < RX_BD_RING_LEN &&
 			     skb_recycle_check(skb,
 				    ugeth->ug_info->uf_info.max_rx_buf_length +

RE: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Wu Jiajun-B06378]

Yes,'if (!skb)' is enough.
You can reproduce transmitting stopping if you use 'if ((bd == ugeth->txBd[txQ])' and run ipforwarding with MTU=64 1Gbps 100%linerate.


-----Original Message-----
From: Anton Vorontsov [mailto:avorontsov@ru.mvista.com] 
Sent: 2010年1月11日 18:53
To: Wu Jiajun-B06378
Cc: linuxppc-dev@ozlabs.org; netdev@vger.kernel.org; lsorense@csclub.uwaterloo.ca; davem@davemloft.net
Subject: Re: [PATCH 1/3] ucc_geth: Fix empty TX queue processing

On Mon, Jan 11, 2010 at 11:47:37AM +0800, Wu Jiajun-B06378 wrote:
>  
> 'bd == ugeth->txBd[txQ]' has two possible statuses: 1)full queue.
> 2)empty queue.
> Removing 'netif_queue_stopped() == 0' will make transmitting stopping 
> when the queue is full.
> 
> I made a new patch for this oops.
[...]
> +		if ((bd == ugeth->txBd[txQ]) && (skb == NULL))
>  			break;

Hm. I wonder why do we need the 'bd == ugeth->txBd[txQ]' check at all? The null skb will cause a kernel oops anyway.

I think the patch below should be sufficient for the fix.
Can you try it? Or if you tell me how to reproduce the issue you observe, I can try it myself.

Thanks a lot!

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c index 41ad2f3..a1a6d06 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3279,13 +3279,12 @@ static int ucc_geth_tx(struct net_device *dev, u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if (bd == ugeth->txBd[txQ]) /* queue empty? */
+		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
+		if (!skb)
 			break;
 
 		dev->stats.tx_packets++;
 
-		skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]];
-
 		if (skb_queue_len(&ugeth->rx_recycle) < RX_BD_RING_LEN &&
 			     skb_recycle_check(skb,
 				    ugeth->ug_info->uf_info.max_rx_buf_length +

[stable] Please apply ucc_geth driver fixes to 2.6.32-stable

[Petri Lehtinen]

(I already sent these patches a few weeks ago and got no reply. I also
sent them to some of the recipients a few minutes ago and failed to
use git-send-email properly. So sorry for the noise.)

Please include the following three patches to 2.6.32-stable. They fix
serious flaws in the ucc_geth ethernet driver, making it unusable (at
least on my system).

They have been included in 2.6.33 and apply cleanly on top of
2.6.32-stable.

[PATCH 1/3] ucc_geth: Fix empty TX queue processing

[Petri Lehtinen]

From: Anton Vorontsov <avorontsov@ru.mvista.com>

Following oops was seen with the ucc_geth driver:

 Unable to handle kernel paging request for data at address 0x00000058
 Faulting instruction address: 0xc024f2fc
 Oops: Kernel access of bad area, sig: 11 [#1]
 [...]
 NIP [c024f2fc] skb_recycle_check+0x14/0x100
 LR [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
 Call Trace:
 [df857d50] [c000b03c] __ipipe_grab_irq+0x3c/0xa4 (unreliable)
 [df857d60] [e30aa0a4] ucc_geth_poll+0xd8/0x4e0 [ucc_geth_driver]
 [df857dd0] [c0258cf8] net_rx_action+0xf8/0x1b8
 [df857e10] [c0032a38] __do_softirq+0xb8/0x13c
 [df857e60] [c00065cc] do_softirq+0xa0/0xac
 [...]

This is because ucc_geth_tx() tries to process an empty queue when
queues are logically stopped. Stopping the queues doesn't disable
polling, and since nowadays ucc_geth_tx() is actually called from
the polling routine, the oops above might pop up.

Fix this by removing 'netif_queue_stopped() == 0' check.

Reported-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Tested-by: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>
Cc: Stable <stable@vger.kernel.org> [2.6.32]
Signed-off-by: David S. Miller <davem@davemloft.net>

(cherry picked from commit 7583605b6d29f1f7f6fc505b883328089f3485ad)
---
 drivers/net/ucc_geth.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 4469f24..20aa5b5 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3273,7 +3273,7 @@ static int ucc_geth_tx(struct net_device *dev, u8 txQ)
 		/* Handle the transmitted buffer and release */
 		/* the BD to be used with the current frame  */
 
-		if ((bd == ugeth->txBd[txQ]) && (netif_queue_stopped(dev) == 0))
+		if (bd == ugeth->txBd[txQ]) /* queue empty? */
 			break;
 
 		dev->stats.tx_packets++;
-- 
1.7.1