[PATCH 0/1] mv643xx_eth: Disable TSO by default

[PATCH 0/1] mv643xx_eth: Disable TSO by default

[Ezequiel Garcia]

Several users ([1], [2]) have been reporting data corruption with TSO on
Kirkwood platforms (i.e. using the mv643xx_eth driver).

Until we manage to find what's causing this, this simple patch will make
the TSO path disabled by default. This patch should be queued for stable,
fixing the TSO feature introduced in v3.16.

The corruption itself is very easy to reproduce: checking md5sum on a mounted
NFS directory gives a different result each time. Same tests using the mvneta
driver (Armada 370/38x/XP SoC) pass with no issues.

Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
are well received.

[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=764162
[2] http://archlinuxarm.org/forum/viewtopic.php?f=9&t=7692

Ezequiel Garcia (1):
  net: mv643xx_eth: Make TSO disabled by default

 drivers/net/ethernet/marvell/mv643xx_eth.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

-- 
2.1.0

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Eric Dumazet]

On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
> Several users ([1], [2]) have been reporting data corruption with TSO on
> Kirkwood platforms (i.e. using the mv643xx_eth driver).
> 
> Until we manage to find what's causing this, this simple patch will make
> the TSO path disabled by default. This patch should be queued for stable,
> fixing the TSO feature introduced in v3.16.
> 
> The corruption itself is very easy to reproduce: checking md5sum on a mounted
> NFS directory gives a different result each time. Same tests using the mvneta
> driver (Armada 370/38x/XP SoC) pass with no issues.
> 
> Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> are well received.

lack of barriers maybe ?

It seems you might need to populate all TX descriptors but delay the
first, like doing the populate in descending order.

If you take a look at txq_submit_skb(), you'll see the final
desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
txq_submit_frag_skb()

You should kick the nick only when all TX descriptors are ready and
committed to memory.

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Thomas Petazzoni]

Dear Eric Dumazet,

On Sat, 01 Nov 2014 10:26:06 -0700, Eric Dumazet wrote:
> On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
> > Several users ([1], [2]) have been reporting data corruption with TSO on
> > Kirkwood platforms (i.e. using the mv643xx_eth driver).
> > 
> > Until we manage to find what's causing this, this simple patch will make
> > the TSO path disabled by default. This patch should be queued for stable,
> > fixing the TSO feature introduced in v3.16.
> > 
> > The corruption itself is very easy to reproduce: checking md5sum on a mounted
> > NFS directory gives a different result each time. Same tests using the mvneta
> > driver (Armada 370/38x/XP SoC) pass with no issues.
> > 
> > Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> > are well received.
> 
> lack of barriers maybe ?
> 
> It seems you might need to populate all TX descriptors but delay the
> first, like doing the populate in descending order.
> 
> If you take a look at txq_submit_skb(), you'll see the final
> desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
> txq_submit_frag_skb()
> 
> You should kick the nick only when all TX descriptors are ready and
> committed to memory.

As far as I know, ARMv5 does not do out-of-order execution, and so on
ARMv5, mb() == rmb() == wmb() == barrier(). But might be a missing
compiler barrier, indeed, as that's not specific to the architecture.

Thanks for the hint!

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Eric Dumazet]

On Sat, 2014-11-01 at 10:26 -0700, Eric Dumazet wrote:
> On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
> > Several users ([1], [2]) have been reporting data corruption with TSO on
> > Kirkwood platforms (i.e. using the mv643xx_eth driver).
> > 
> > Until we manage to find what's causing this, this simple patch will make
> > the TSO path disabled by default. This patch should be queued for stable,
> > fixing the TSO feature introduced in v3.16.
> > 
> > The corruption itself is very easy to reproduce: checking md5sum on a mounted
> > NFS directory gives a different result each time. Same tests using the mvneta
> > driver (Armada 370/38x/XP SoC) pass with no issues.
> > 
> > Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> > are well received.
> 
> lack of barriers maybe ?
> 
> It seems you might need to populate all TX descriptors but delay the
> first, like doing the populate in descending order.
> 
> If you take a look at txq_submit_skb(), you'll see the final
> desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
> txq_submit_frag_skb()
> 
> You should kick the nick only when all TX descriptors are ready and
> committed to memory.
> 

Untested patch would be :

diff --git a/drivers/net/ethernet/marvell/mv643xx_eth.c b/drivers/net/ethernet/marvell/mv643xx_eth.c
index b151a949f352a20ec8e74b4f3a7b6bb194ce841c..44789cc9a263992f91e46006d7e12703a2824cb4 100644
--- a/drivers/net/ethernet/marvell/mv643xx_eth.c
+++ b/drivers/net/ethernet/marvell/mv643xx_eth.c
@@ -773,7 +773,8 @@ txq_put_data_tso(struct net_device *dev, struct tx_queue *txq,
 }
 
 static inline void
-txq_put_hdr_tso(struct sk_buff *skb, struct tx_queue *txq, int length)
+txq_put_hdr_tso(struct sk_buff *skb, struct tx_queue *txq, int length,
+		struct tx_desc **pdesc, u32 *cmd)
 {
 	struct mv643xx_eth_private *mp = txq_to_mp(txq);
 	int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
@@ -797,9 +798,13 @@ txq_put_hdr_tso(struct sk_buff *skb, struct tx_queue *txq, int length)
 	desc->byte_cnt = hdr_len;
 	desc->buf_ptr = txq->tso_hdrs_dma +
 			txq->tx_curr_desc * TSO_HEADER_SIZE;
-	desc->cmd_sts = cmd_csum | BUFFER_OWNED_BY_DMA  | TX_FIRST_DESC |
-				   GEN_CRC;
-
+	cmd_csum |= BUFFER_OWNED_BY_DMA  | TX_FIRST_DESC | GEN_CRC;
+	if (*pdesc == NULL) {
+		*pdesc = desc;
+		*cmd = cmd_csum;
+	} else {
+		desc->cmd_sts = cmd_csum;
+	}
 	txq->tx_curr_desc++;
 	if (txq->tx_curr_desc == txq->tx_ring_size)
 		txq->tx_curr_desc = 0;
@@ -813,6 +818,8 @@ static int txq_submit_tso(struct tx_queue *txq, struct sk_buff *skb,
 	int desc_count = 0;
 	struct tso_t tso;
 	int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
+	struct tx_desc *desc = NULL;
+	u32 cmd_sts = 0;
 
 	/* Count needed descriptors */
 	if ((txq->tx_desc_count + tso_count_descs(skb)) >= txq->tx_ring_size) {
@@ -834,7 +841,7 @@ static int txq_submit_tso(struct tx_queue *txq, struct sk_buff *skb,
 		/* prepare packet headers: MAC + IP + TCP */
 		hdr = txq->tso_hdrs + txq->tx_curr_desc * TSO_HEADER_SIZE;
 		tso_build_hdr(skb, hdr, &tso, data_left, total_len == 0);
-		txq_put_hdr_tso(skb, txq, data_left);
+		txq_put_hdr_tso(skb, txq, data_left, &desc, &cmd_sts);
 
 		while (data_left > 0) {
 			int size;
@@ -854,6 +861,10 @@ static int txq_submit_tso(struct tx_queue *txq, struct sk_buff *skb,
 	__skb_queue_tail(&txq->tx_skb, skb);
 	skb_tx_timestamp(skb);
 
+	/* ensure all other descriptors are written before first cmd_sts */
+	wmb();
+	desc->cmd_sts = cmd_sts;
+
 	/* clear TX_END status */
 	mp->work_tx_end &= ~(1 << txq->index);
 

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Sat, 01 Nov 2014 10:26:06 -0700

> On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
>> Several users ([1], [2]) have been reporting data corruption with TSO on
>> Kirkwood platforms (i.e. using the mv643xx_eth driver).
>> 
>> Until we manage to find what's causing this, this simple patch will make
>> the TSO path disabled by default. This patch should be queued for stable,
>> fixing the TSO feature introduced in v3.16.
>> 
>> The corruption itself is very easy to reproduce: checking md5sum on a mounted
>> NFS directory gives a different result each time. Same tests using the mvneta
>> driver (Armada 370/38x/XP SoC) pass with no issues.
>> 
>> Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
>> are well received.
> 
> lack of barriers maybe ?
> 
> It seems you might need to populate all TX descriptors but delay the
> first, like doing the populate in descending order.
> 
> If you take a look at txq_submit_skb(), you'll see the final
> desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
> txq_submit_frag_skb()
> 
> You should kick the nick only when all TX descriptors are ready and
> committed to memory.

Yes, please look into whether doing something like this fixes the
problem.

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Eric Dumazet]

On Sat, 2014-11-01 at 18:33 +0100, Thomas Petazzoni wrote:
> Dear Eric Dumazet,

You can call me Eric ;)

> As far as I know, ARMv5 does not do out-of-order execution, and so on
> ARMv5, mb() == rmb() == wmb() == barrier(). But might be a missing
> compiler barrier, indeed, as that's not specific to the architecture.
> 
> Thanks for the hint!

I think the wmb() is needed, but the more worrying problem is the NIC
can start transmitting a segment while the data payload is not yet
placed in TX descriptor.

Normally, the patch I cooked should avoid this.

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Ezequiel Garcia]

[-- Attachment #1: Type: text/plain, Size: 1716 bytes --]

On 11/01/2014 02:37 PM, Eric Dumazet wrote:
> On Sat, 2014-11-01 at 10:26 -0700, Eric Dumazet wrote:
>> On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
>>> Several users ([1], [2]) have been reporting data corruption with TSO on
>>> Kirkwood platforms (i.e. using the mv643xx_eth driver).
>>>
>>> Until we manage to find what's causing this, this simple patch will make
>>> the TSO path disabled by default. This patch should be queued for stable,
>>> fixing the TSO feature introduced in v3.16.
>>>
>>> The corruption itself is very easy to reproduce: checking md5sum on a mounted
>>> NFS directory gives a different result each time. Same tests using the mvneta
>>> driver (Armada 370/38x/XP SoC) pass with no issues.
>>>
>>> Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
>>> are well received.
>>
>> lack of barriers maybe ?
>>

Yup, that was my initial thought as well...

>> It seems you might need to populate all TX descriptors but delay the
>> first, like doing the populate in descending order.
>>
>> If you take a look at txq_submit_skb(), you'll see the final
>> desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
>> txq_submit_frag_skb()
>>
>> You should kick the nick only when all TX descriptors are ready and
>> committed to memory.
>>
> 
> Untested patch would be :
> 

Yeah, it makes sense. I'm still seeing the corruption after applying
your patch.

However, maybe we are onto something. I'll see about taking a closer
look and give this some more thought.

Thanks for the hint!
-- 
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

RE: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[David Laight]

From: Eric Dumazet
> On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
> > Several users ([1], [2]) have been reporting data corruption with TSO on
> > Kirkwood platforms (i.e. using the mv643xx_eth driver).
> >
> > Until we manage to find what's causing this, this simple patch will make
> > the TSO path disabled by default. This patch should be queued for stable,
> > fixing the TSO feature introduced in v3.16.
> >
> > The corruption itself is very easy to reproduce: checking md5sum on a mounted
> > NFS directory gives a different result each time. Same tests using the mvneta
> > driver (Armada 370/38x/XP SoC) pass with no issues.
> >
> > Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> > are well received.
> 
> lack of barriers maybe ?
> 
> It seems you might need to populate all TX descriptors but delay the
> first, like doing the populate in descending order.
> 
> If you take a look at txq_submit_skb(), you'll see the final
> desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
> txq_submit_frag_skb()
> 
> You should kick the nick only when all TX descriptors are ready and
> committed to memory.

Don't forget that the nick might process the first descriptor without
being given a 'kick' - it will read it when it finishes processing the
previous frame.
This also means that you have to be careful about the order of the writes
to the first descriptor.

	David

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Eric Dumazet]

On Mon, 2014-11-03 at 14:51 +0000, David Laight wrote:
> From: Eric Dumazet
> > On Sat, 2014-11-01 at 12:30 -0300, Ezequiel Garcia wrote:
> > > Several users ([1], [2]) have been reporting data corruption with TSO on
> > > Kirkwood platforms (i.e. using the mv643xx_eth driver).
> > >
> > > Until we manage to find what's causing this, this simple patch will make
> > > the TSO path disabled by default. This patch should be queued for stable,
> > > fixing the TSO feature introduced in v3.16.
> > >
> > > The corruption itself is very easy to reproduce: checking md5sum on a mounted
> > > NFS directory gives a different result each time. Same tests using the mvneta
> > > driver (Armada 370/38x/XP SoC) pass with no issues.
> > >
> > > Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> > > are well received.
> > 
> > lack of barriers maybe ?
> > 
> > It seems you might need to populate all TX descriptors but delay the
> > first, like doing the populate in descending order.
> > 
> > If you take a look at txq_submit_skb(), you'll see the final
> > desc->cmd_sts = cmd_sts (line 959) is done _after_ frags were cooked by
> > txq_submit_frag_skb()
> > 
> > You should kick the nick only when all TX descriptors are ready and
> > committed to memory.
> 
> Don't forget that the nick might process the first descriptor without
> being given a 'kick' - it will read it when it finishes processing the
> previous frame.
> This also means that you have to be careful about the order of the writes
> to the first descriptor.

This is what I implied and implemented in the patch ;)

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Karl Beldan]

On Sat, Nov 01, 2014 at 12:30:19PM -0300, Ezequiel Garcia wrote:
> Several users ([1], [2]) have been reporting data corruption with TSO on
> Kirkwood platforms (i.e. using the mv643xx_eth driver).
> 
> Until we manage to find what's causing this, this simple patch will make
> the TSO path disabled by default. This patch should be queued for stable,
> fixing the TSO feature introduced in v3.16.
> 
> The corruption itself is very easy to reproduce: checking md5sum on a mounted
> NFS directory gives a different result each time. Same tests using the mvneta
> driver (Armada 370/38x/XP SoC) pass with no issues.
> 
> Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> are well received.
> 

Hi,

Can you try this :

@@ -1067,7 +1082,8 @@ static int txq_reclaim(struct tx_queue *txq, int budget, int force)
 		txq->tx_desc_count--;
 
 		skb = NULL;
-		if (cmd_sts & TX_LAST_DESC)
+		if ((cmd_sts & (TX_LAST_DESC | TX_ENABLE_INTERRUPT)) ==
+			       (TX_LAST_DESC | TX_ENABLE_INTERRUPT))
 			skb = __skb_dequeue(&txq->tx_skb);
 
 		if (cmd_sts & ERROR_SUMMARY) {

-- 
Karl

Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

[Karl Beldan]

On Wed, Nov 05, 2014 at 08:39:26AM +0000, Ian Campbell wrote:
> On Tue, 2014-11-04 at 15:20 +0100, Karl Beldan wrote:
> > On Sat, Nov 01, 2014 at 12:30:19PM -0300, Ezequiel Garcia wrote:
> > > Several users ([1], [2]) have been reporting data corruption with TSO on
> > > Kirkwood platforms (i.e. using the mv643xx_eth driver).
> > > 
> > > Until we manage to find what's causing this, this simple patch will make
> > > the TSO path disabled by default. This patch should be queued for stable,
> > > fixing the TSO feature introduced in v3.16.
> > > 
> > > The corruption itself is very easy to reproduce: checking md5sum on a mounted
> > > NFS directory gives a different result each time. Same tests using the mvneta
> > > driver (Armada 370/38x/XP SoC) pass with no issues.
> > > 
> > > Frankly, I'm a bit puzzled about this, and so any ideas or debugging hints
> > > are well received.
> > > 
> > 
> > Hi,
> > 
> > Can you try this :
> 
> It fixes things for me, thanks!
> 
> Tested-by: Ian Campbell <ijc@hellion.org.uk>
> 

Good thing, thanks for your feedbak Ian !
 
Karl

[PATCH 0/1] mv643xx_eth: Disable TSO by default

[Dr. Uwe Meyer-Gruhl]

> On Wed, Nov 05, 2014 at 08:39:26AM +0000, Ian Campbell wrote:
>> On Tue, 2014-11-04 at 15:20 +0100, Karl Beldan wrote:
>>> On Sat, Nov 01, 2014 at 12:30:19PM -0300, Ezequiel Garcia wrote:
>>>> Several users ([1], [2]) have been reporting data corruption
>>>> with TSO on Kirkwood platforms (i.e. using the mv643xx_eth
>>>> driver).
>>>>
>>>> Until we manage to find what's causing this, this simple patch  will make
>>>> the TSO path disabled by default. This patch should be queued
for stable,
>>>> fixing the TSO feature introduced in v3.16.
>>>>
>>>> The corruption itself is very easy to reproduce: checkingmd5sum on a mounted
>>>> NFS directory gives a different result each time. Same tests  using the mvneta
>>>> driver (Armada 370/38x/XP SoC) pass with no issues.
>>>>
>>>> Frankly, I'm a bit puzzled about this, and so any ideas ordebugging hints
>>>> are well received.
>>>>
>>>
>>> Hi,
>>>
>>> Can you try this :
>>
>> It fixes things for me, thanks!
>>
>> Tested-by: Ian Campbell <ijc@hellion.org.uk>
>>
>
> Good thing, thanks for your feedbak Ian !
>
> Karl --

That would be a good thing - although: Neither the patch to disable TSO
altogether nor the one that fixes the underlying problem actually made
it to the official kernel source tree, so it is still present in all 
kernels > 3.16 - I just stumbled over this in the current 4.0.4 version.

The fixes in the thread
http://marc.info/?l=linux-netdev&m=141517941900547&w=2 are not
applicable any more to the current driver from the 4.0 kernel, as the
whole respective logic seems to have been changed meanwhile, sadly 
without fixing the problem. Disabling TSO completely still works, though.

Can someone in the know please suggest a working fix to the kernel
maintainers, preferably one that does not resort to disable TSO?


Uwe
--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3740 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20150529/a14a7b4f/attachment.p7s>

[PATCH 0/1] mv643xx_eth: Disable TSO by default

[Ezequiel Garcia]

On 05/29/2015 03:16 AM, Dr. Uwe Meyer-Gruhl wrote:
>> On Wed, Nov 05, 2014 at 08:39:26AM +0000, Ian Campbell wrote:
>>> On Tue, 2014-11-04 at 15:20 +0100, Karl Beldan wrote:
>>>> On Sat, Nov 01, 2014 at 12:30:19PM -0300, Ezequiel Garcia wrote:
>>>>> Several users ([1], [2]) have been reporting data corruption
>>>>> with TSO on Kirkwood platforms (i.e. using the mv643xx_eth
>>>>> driver).
>>>>>
>>>>> Until we manage to find what's causing this, this simple patch 
>>>>> will make
>>>>> the TSO path disabled by default. This patch should be queued
> for stable,
>>>>> fixing the TSO feature introduced in v3.16.
>>>>>
>>>>> The corruption itself is very easy to reproduce: checkingmd5sum on
>>>>> a mounted
>>>>> NFS directory gives a different result each time. Same tests  using
>>>>> the mvneta
>>>>> driver (Armada 370/38x/XP SoC) pass with no issues.
>>>>>
>>>>> Frankly, I'm a bit puzzled about this, and so any ideas ordebugging
>>>>> hints
>>>>> are well received.
>>>>>
>>>>
>>>> Hi,
>>>>
>>>> Can you try this :
>>>
>>> It fixes things for me, thanks!
>>>
>>> Tested-by: Ian Campbell <ijc@hellion.org.uk>
>>>
>>
>> Good thing, thanks for your feedbak Ian !
>>
>> Karl --
> 
> That would be a good thing - although: Neither the patch to disable TSO
> altogether nor the one that fixes the underlying problem actually made
> it to the official kernel source tree, so it is still present in all
> kernels > 3.16 - I just stumbled over this in the current 4.0.4 version.
> 
> The fixes in the thread
> http://marc.info/?l=linux-netdev&m=141517941900547&w=2 are not
> applicable any more to the current driver from the 4.0 kernel, as the
> whole respective logic seems to have been changed meanwhile, sadly
> without fixing the problem. Disabling TSO completely still works, though.
> 
> Can someone in the know please suggest a working fix to the kernel
> maintainers, preferably one that does not resort to disable TSO?
> 

Hello Uwe,

Thanks for reporting. There wasn't any patch left behind. The above fix
was merged (and applied to the stable tree) as this commit:

commit 2c2a9cbd64387d6b70ac5db013e9bfe9412c7354
Author: Karl Beldan <karl.beldan@rivierawaves.com>
Date:   Wed Nov 5 15:32:59 2014 +0100

    net: mv643xx_eth: reclaim TX skbs only when released by the HW

Can you explain in detail what sort of problem are you running into,
which kernel version are you using, etc.?
-- 
Ezequiel Garcia, VanguardiaSur
www.vanguardiasur.com.ar

[PATCH 0/1] mv643xx_eth: Disable TSO by default

[Dr. Uwe Meyer-Gruhl]

Am 29.05.2015 um 15:07 schrieb Ezequiel Garcia:
> On 05/29/2015 03:16 AM, Dr. Uwe Meyer-Gruhl wrote:
>>> On Wed, Nov 05, 2014 at 08:39:26AM +0000, Ian Campbell wrote:
>>>> On Tue, 2014-11-04 at 15:20 +0100, Karl Beldan wrote:
>>>>> On Sat, Nov 01, 2014 at 12:30:19PM -0300, Ezequiel Garcia wrote:
>>>>>> Several users ([1], [2]) have been reporting data corruption
>>>>>> with TSO on Kirkwood platforms (i.e. using the mv643xx_eth
>>>>>> driver).
>>>>>>
>>>>>> Until we manage to find what's causing this, this simple patch
>>>>>> will make
>>>>>> the TSO path disabled by default. This patch should be queued
>> for stable,
>>>>>> fixing the TSO feature introduced in v3.16.
>>>>>>
>>>>>> The corruption itself is very easy to reproduce: checkingmd5sum on
>>>>>> a mounted
>>>>>> NFS directory gives a different result each time. Same tests  using
>>>>>> the mvneta
>>>>>> driver (Armada 370/38x/XP SoC) pass with no issues.
>>>>>>
>>>>>> Frankly, I'm a bit puzzled about this, and so any ideas ordebugging
>>>>>> hints
>>>>>> are well received.
>>>>>>
>>>>> Hi,
>>>>>
>>>>> Can you try this :
>>>> It fixes things for me, thanks!
>>>>
>>>> Tested-by: Ian Campbell <ijc@hellion.org.uk>
>>>>
>>> Good thing, thanks for your feedbak Ian !
>>>
>>> Karl --
>> That would be a good thing - although: Neither the patch to disable TSO
>> altogether nor the one that fixes the underlying problem actually made
>> it to the official kernel source tree, so it is still present in all
>> kernels > 3.16 - I just stumbled over this in the current 4.0.4 version.
>>
>> The fixes in the thread
>> http://marc.info/?l=linux-netdev&m=141517941900547&w=2 are not
>> applicable any more to the current driver from the 4.0 kernel, as the
>> whole respective logic seems to have been changed meanwhile, sadly
>> without fixing the problem. Disabling TSO completely still works, though.
>>
>> Can someone in the know please suggest a working fix to the kernel
>> maintainers, preferably one that does not resort to disable TSO?
>>
> Hello Uwe,
>
> Thanks for reporting. There wasn't any patch left behind. The above fix
> was merged (and applied to the stable tree) as this commit:
>
> commit 2c2a9cbd64387d6b70ac5db013e9bfe9412c7354
> Author: Karl Beldan <karl.beldan@rivierawaves.com>
> Date:   Wed Nov 5 15:32:59 2014 +0100
>
>      net: mv643xx_eth: reclaim TX skbs only when released by the HW
>
> Can you explain in detail what sort of problem are you running into,
> which kernel version are you using, etc.?
Sure, the same kind as had been reported initially.

Background: I have done a Debian port running on an Iomega iConnect (a 
Kirkwood variant).

One of my users reported corrupted data when he used the machine as a 
file and media server over Samba and DLNA. He ruled out any hardware or 
other faults and pointed to my kernel(s). He actually tried versions 
3.18.1 and 4.0.4. However, my kernels are mostly plain vanilla from 
kernel.org without any relevant patches to the code itself. The user 
also found that a kernel 3.14 ran O.K. in the same context.

When I tried, I could just use md5sum on the /usr subtree repeatedly 
without problems locally, but over NFS with the iConnect as a server, I 
had differences on every try in different files each time.

After ruling out network issues as well, it was clear that the NIC 
driver must be the culprit and after a bit of searching which one it 
actually was, I quickly found the messages about the broken TSO feature 
in mv643xx_eth. Then, I tried the suggested fix of "ethtool -K eth0 tso 
off" and voila - the errors are gone. For now, I use a patch to change 
the default to TSO off, but an upstream patch to handle TSO correctly 
would be much better.


Uwe
-- 


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3740 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20150529/12a23844/attachment.p7s>