KS8695: possible NAPI issue

KS8695: possible NAPI issue

[Yegor Yefremov]

I'm using 2.6.33 kernel and I noticed such a strange behavior:

after system start I transfer one file via netcat from my development
host, after this transfer the network is not functioning i.e. no pings
possible etc.

To narrow down the problem I checked out this commit
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
Till here the network driver is functioning as intended, but after
NAPI introduction I have this issue. With latest git-pull of "Linus'
kernel tree" I can't even ping right after the systems start.

Any Ideas? What am I missing?

Regards,
Yegor

Re: KS8695: possible NAPI issue

[Dick Hollenbeck]

Yegor Yefremov wrote:
> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>
> after system start I transfer one file via netcat from my development
> host, after this transfer the network is not functioning i.e. no pings
> possible etc.
>
> To narrow down the problem I checked out this commit
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
> Till here the network driver is functioning as intended, but after
> NAPI introduction I have this issue. With latest git-pull of "Linus'
> kernel tree" I can't even ping right after the systems start.
>
> Any Ideas? What am I missing?
>   

No idea, although I am using the same ARM chip, my kernel is at 
2.6.30.5, and except for this occasional loss of connection I get, the 
ethernet driver seems to work better than what you are reporting.

I would get your printk() working to the internal ring buffer, and then 
simply instrument the key pieces of code with printk()s until the 
situation becomes clear.  You can also watch

# cat /proc/interrupts to see which interrupt handlers are firing, 
before and after an activity.  Reception of a frame all starts with an 
interrupt handler usually.

printk() should be able to print even from within an interrupt handler, 
so long as you are only going to that internal ring buffer.


Dick

> Regards,
> Yegor
>
>   

Re: KS8695: possible NAPI issue

[figo zhang]

2010/3/2 Dick Hollenbeck <dick@softplc.com>:
> Yegor Yefremov wrote:
>>
>> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>>
>> after system start I transfer one file via netcat from my development
>> host, after this transfer the network is not functioning i.e. no pings
>> possible etc.
>>
>> To narrow down the problem I checked out this commit
>>
>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
>> Till here the network driver is functioning as intended, but after
>> NAPI introduction I have this issue. With latest git-pull of "Linus'
>> kernel tree" I can't even ping right after the systems start.
>>
>> Any Ideas? What am I missing?
>>
>
> No idea, although I am using the same ARM chip, my kernel is at 2.6.30.5,
> and except for this occasional loss of connection I get, the ethernet driver
> seems to work better than what you are reporting.

from linux-2.6.32, this ethernet  driver have add NAPI support, would
you like to
try using this version?

Re: KS8695: possible NAPI issue

[Dick Hollenbeck]

figo zhang wrote:
> 2010/3/2 Dick Hollenbeck <dick@softplc.com>:
>   
>> Yegor Yefremov wrote:
>>     
>>> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>>>
>>> after system start I transfer one file via netcat from my development
>>> host, after this transfer the network is not functioning i.e. no pings
>>> possible etc.
>>>
>>> To narrow down the problem I checked out this commit
>>>
>>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
>>> Till here the network driver is functioning as intended, but after
>>> NAPI introduction I have this issue. With latest git-pull of "Linus'
>>> kernel tree" I can't even ping right after the systems start.
>>>
>>> Any Ideas? What am I missing?
>>>
>>>       
>> No idea, although I am using the same ARM chip, my kernel is at 2.6.30.5,
>> and except for this occasional loss of connection I get, the ethernet driver
>> seems to work better than what you are reporting.
>>     
>
> from linux-2.6.32, this ethernet  driver have add NAPI support, would
> you like to
> try using this version?
>   

Seems like a lot of work for me, given that my 2.6.30.5 kernel is mostly 
working.   Yegor is having the problems on the newer kernels, newer than 
mine.  Sounds like a warning flag to me, not an incentive.

What is NAPI and why do I care?

Dick

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

>>>> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>>>>
>>>> after system start I transfer one file via netcat from my development
>>>> host, after this transfer the network is not functioning i.e. no pings
>>>> possible etc.
>>>>
>>>> To narrow down the problem I checked out this commit
>>>>
>>>>
>>>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
>>>> Till here the network driver is functioning as intended, but after
>>>> NAPI introduction I have this issue. With latest git-pull of "Linus'
>>>> kernel tree" I can't even ping right after the systems start.
>>>>
>>>> Any Ideas? What am I missing?
>>>>
>>>>
>>>
>>> No idea, although I am using the same ARM chip, my kernel is at 2.6.30.5,
>>> and except for this occasional loss of connection I get, the ethernet
>>> driver
>>> seems to work better than what you are reporting.
>>>
>>
>> from linux-2.6.32, this ethernet  driver have add NAPI support, would
>> you like to
>> try using this version?
>>
>
> Seems like a lot of work for me, given that my 2.6.30.5 kernel is mostly
> working.   Yegor is having the problems on the newer kernels, newer than
> mine.  Sounds like a warning flag to me, not an incentive.
>
> What is NAPI and why do I care?

Here is Wikipedia article: http://en.wikipedia.org/wiki/NAPI

The problem also is that I really need a kernel version not less than
2.6.33 because of the support of other platforms.

I sniffed a little with wireshark and the symptoms are the following:

after netcat transfer the system sends only ARP requests and no other packets.

I just git-pulled the latest Linus tree and the network still has the
above described issues.

Figo, could you test the latest kernel with netcat transfers? So you
could reproduce this behavior?

Yegor

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

>>> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>>>
>>> after system start I transfer one file via netcat from my development
>>> host, after this transfer the network is not functioning i.e. no pings
>>> possible etc.
>>>
>>> To narrow down the problem I checked out this commit
>>>
>>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
>>> Till here the network driver is functioning as intended, but after
>>> NAPI introduction I have this issue. With latest git-pull of "Linus'
>>> kernel tree" I can't even ping right after the systems start.
>>>
>>> Any Ideas? What am I missing?
>>>
>>
>> No idea, although I am using the same ARM chip, my kernel is at 2.6.30.5,
>> and except for this occasional loss of connection I get, the ethernet driver
>> seems to work better than what you are reporting.
>
> from linux-2.6.32, this ethernet  driver have add NAPI support, would
> you like to
> try using this version?

Figo, I just looked at the 2.6.32 kernel and found out that it
doesn't' contain NAPI support (see the history
http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.32.y.git;a=history;f=drivers/net/arm/ks8695net.c;h=2a7b7745cc55b54cefada690b0ea47865b54b6b9;hb=HEAD).
So 2.6.33 is the first stable kernel to support NAPI.

I made some further tests but I can't find the reason for network
failures: the interrupts are coming and the packets will be received
and sent so that rx/tx counters would be increased, but none of them
seem to reach the IP stack. I don't know where and why the things get
desynchronized.

Could you try the 2.6.33 or Linus tree? Thank you in advance.

Regards,
Yegor

Re: KS8695: possible NAPI issue

[figo zhang]

2010/3/4 Yegor Yefremov <yegorslists@googlemail.com>:
>>>> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>>>>
>>>> after system start I transfer one file via netcat from my development
>>>> host, after this transfer the network is not functioning i.e. no pings
>>>> possible etc.
>>>>
>>>> To narrow down the problem I checked out this commit
>>>>
>>>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
>>>> Till here the network driver is functioning as intended, but after
>>>> NAPI introduction I have this issue. With latest git-pull of "Linus'
>>>> kernel tree" I can't even ping right after the systems start.
>>>>
>>>> Any Ideas? What am I missing?
>>>>
>>>
>>> No idea, although I am using the same ARM chip, my kernel is at 2.6.30.5,
>>> and except for this occasional loss of connection I get, the ethernet driver
>>> seems to work better than what you are reporting.
>>
>> from linux-2.6.32, this ethernet  driver have add NAPI support, would
>> you like to
>> try using this version?
>
> Figo, I just looked at the 2.6.32 kernel and found out that it
> doesn't' contain NAPI support (see the history
> http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.32.y.git;a=history;f=drivers/net/arm/ks8695net.c;h=2a7b7745cc55b54cefada690b0ea47865b54b6b9;hb=HEAD).
> So 2.6.33 is the first stable kernel to support NAPI.
>
> I made some further tests but I can't find the reason for network
> failures: the interrupts are coming and the packets will be received
> and sent so that rx/tx counters would be increased, but none of them
> seem to reach the IP stack. I don't know where and why the things get
> desynchronized.
>
> Could you try the 2.6.33 or Linus tree? Thank you in advance.

yes, i have test it on linux-2.6.33, it is work well.

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

On Fri, Mar 5, 2010 at 7:54 AM, figo zhang <figo1802@gmail.com> wrote:
> 2010/3/4 Yegor Yefremov <yegorslists@googlemail.com>:
>>>>> I'm using 2.6.33 kernel and I noticed such a strange behavior:
>>>>>
>>>>> after system start I transfer one file via netcat from my development
>>>>> host, after this transfer the network is not functioning i.e. no pings
>>>>> possible etc.
>>>>>
>>>>> To narrow down the problem I checked out this commit
>>>>>
>>>>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=451f14439847db302e5104c44458b2dbb4b1829d.
>>>>> Till here the network driver is functioning as intended, but after
>>>>> NAPI introduction I have this issue. With latest git-pull of "Linus'
>>>>> kernel tree" I can't even ping right after the systems start.
>>>>>
>>>>> Any Ideas? What am I missing?
>>>>>
>>>>
>>>> No idea, although I am using the same ARM chip, my kernel is at 2.6.30.5,
>>>> and except for this occasional loss of connection I get, the ethernet driver
>>>> seems to work better than what you are reporting.
>>>
>>> from linux-2.6.32, this ethernet  driver have add NAPI support, would
>>> you like to
>>> try using this version?
>>
>> Figo, I just looked at the 2.6.32 kernel and found out that it
>> doesn't' contain NAPI support (see the history
>> http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.32.y.git;a=history;f=drivers/net/arm/ks8695net.c;h=2a7b7745cc55b54cefada690b0ea47865b54b6b9;hb=HEAD).
>> So 2.6.33 is the first stable kernel to support NAPI.
>>
>> I made some further tests but I can't find the reason for network
>> failures: the interrupts are coming and the packets will be received
>> and sent so that rx/tx counters would be increased, but none of them
>> seem to reach the IP stack. I don't know where and why the things get
>> desynchronized.
>>
>> Could you try the 2.6.33 or Linus tree? Thank you in advance.
>
> yes, i have test it on linux-2.6.33, it is work well.

Thanks for testing. What kind of tests have you made? And with what data amount?

My tests look like following:

1. system start
2. ping one host: O.K.
3. nc -l -p 5000 > /var/test (about 1Mb): O.K.
4. ping the same host: failed

and even if the fourth test succeeds, if I then send a file from other
host the next ping or any other communication fails. I just tried to
make some SSH sessions. It goes without a problem till I issue a ping
after that.

If I test the system with nuttcp I have no problem.

Do you have any idea how to debug this? If I remove the NAPI support I
have no problem with the same kernel and environment. I'd really like
to get the whole stuff working.

Regards,
Yegor

Re: KS8695: possible NAPI issue

[Figo.zhang]

> 
> and even if the fourth test succeeds, if I then send a file from other
> host the next ping or any other communication fails. I just tried to
> make some SSH sessions. It goes without a problem till I issue a ping
> after that.
> 
> If I test the system with nuttcp I have no problem.
the network driver seems work well for system. the "netcat" application
is build in busybox? you can check busybox configure. 
> 
> Do you have any idea how to debug this? If I remove the NAPI support I
> have no problem with the same kernel and environment. I'd really like
> to get the whole stuff working.
> 
> Regards,
> Yegor

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

On Fri, Mar 5, 2010 at 2:52 PM, Figo.zhang <figo1802@gmail.com> wrote:
>
>>
>> and even if the fourth test succeeds, if I then send a file from other
>> host the next ping or any other communication fails. I just tried to
>> make some SSH sessions. It goes without a problem till I issue a ping
>> after that.
>>
>> If I test the system with nuttcp I have no problem.
> the network driver seems work well for system. the "netcat" application
> is build in busybox? you can check busybox configure.

I'm using Debian for ARM. So it is standard netcat for Debian 5.0.

Re: KS8695: possible NAPI issue

[Figo.zhang]

> 
> My tests look like following:
> 
> 1. system start
> 2. ping one host: O.K.
> 3. nc -l -p 5000 > /var/test (about 1Mb): O.K.
> 4. ping the same host: failed
1. type "arp" command, see the arp is exit or expire?
2. at this point, see is it still have RX interrpt and receive packet in
ks8695_rx()? (in some watchpoint add printk). 

> 
> and even if the fourth test succeeds, if I then send a file from other
> host the next ping or any other communication fails. I just tried to
> make some SSH sessions. It goes without a problem till I issue a ping
> after that.
> 
> If I test the system with nuttcp I have no problem.
> 
> Do you have any idea how to debug this? If I remove the NAPI support I
> have no problem with the same kernel and environment. I'd really like
> to get the whole stuff working.
> 
> Regards,
> Yegor

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

>> My tests look like following:
>>
>> 1. system start
>> 2. ping one host: O.K.
>> 3. nc -l -p 5000 > /var/test (about 1Mb): O.K.
>> 4. ping the same host: failed
> 1. type "arp" command, see the arp is exit or expire?

debian:~# nc -l -p 5000 > /var/zImage

debian:~# ping -c 1 192.168.1.38

PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.

>From 192.168.1.66 icmp_seq=1 Destination Host Unreachable



--- 192.168.1.38 ping statistics ---

1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms



debian:~# arp

Address                  HWtype  HWaddress           Flags Mask            Iface
192.168.1.38          (incomplete)                              eth0

192.168.1.36             ether   00:10:18:39:19:aa   C                     eth0


> 2. at this point, see is it still have RX interrpt and receive packet in
> ks8695_rx()? (in some watchpoint add printk).

I've inserted some printks and I can see that interrupts are coming
and the received count will be increased. I can also see my system
sending arp requests. Even this ping request is visible in wireshark
and its reply, but the ping utility sees nothing of it.

Re: KS8695: possible NAPI issue

[Stephen Hemminger]

On Fri, 5 Mar 2010 16:00:11 +0100
Yegor Yefremov <yegorslists@googlemail.com> wrote:

> >> My tests look like following:
> >>
> >> 1. system start
> >> 2. ping one host: O.K.
> >> 3. nc -l -p 5000 > /var/test (about 1Mb): O.K.
> >> 4. ping the same host: failed
> > 1. type "arp" command, see the arp is exit or expire?
> 
> debian:~# nc -l -p 5000 > /var/zImage
> 
> debian:~# ping -c 1 192.168.1.38
> 
> PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.
> 
> From 192.168.1.66 icmp_seq=1 Destination Host Unreachable
> 
> 
> 
> --- 192.168.1.38 ping statistics ---
> 
> 1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
> 
> 
> 
> debian:~# arp
> 
> Address                  HWtype  HWaddress           Flags Mask            Iface
> 192.168.1.38          (incomplete)                              eth0
> 
> 192.168.1.36             ether   00:10:18:39:19:aa   C                     eth0
> 
> 
> > 2. at this point, see is it still have RX interrpt and receive packet in
> > ks8695_rx()? (in some watchpoint add printk).
> 
> I've inserted some printks and I can see that interrupts are coming
> and the received count will be increased. I can also see my system
> sending arp requests. Even this ping request is visible in wireshark
> and its reply, but the ping utility sees nothing of it.


One of the requirements for NAPI to work is that the IRQ is level triggered.
Otherwise there can be a race between packet arrival and the end of
NAPI processing:

static int ks8695_poll(struct napi_struct *napi, int budget)
{
	struct ks8695_priv *ksp = container_of(napi, struct ks8695_priv, napi);
	unsigned long  work_done;

	unsigned long isr = readl(KS8695_IRQ_VA + KS8695_INTEN);
	unsigned long mask_bit = 1 << ks8695_get_rx_enable_bit(ksp);

	work_done = ks8695_rx(ksp, budget);

	if (work_done < budget) {
		unsigned long flags;
		spin_lock_irqsave(&ksp->rx_lock, flags);
>>> packet arrives >>
		/*enable rx interrupt*/
		writel(isr | mask_bit, KS8695_IRQ_VA + KS8695_INTEN);
>>> or here >>
		__napi_complete(napi);
		spin_unlock_irqrestore(&ksp->rx_lock, flags);

If the hardware can't be reprogrammed to level mode;
the solution used in some drivers is to poll for lost packet after re-enabling
NAPI.



-- 

Re: KS8695: possible NAPI issue

[Figo.zhang]

> 
> One of the requirements for NAPI to work is that the IRQ is level triggered.
> Otherwise there can be a race between packet arrival and the end of
> NAPI processing:
> 
> static int ks8695_poll(struct napi_struct *napi, int budget)
> {
> 	struct ks8695_priv *ksp = container_of(napi, struct ks8695_priv, napi);
> 	unsigned long  work_done;
> 
> 	unsigned long isr = readl(KS8695_IRQ_VA + KS8695_INTEN);
> 	unsigned long mask_bit = 1 << ks8695_get_rx_enable_bit(ksp);
> 
> 	work_done = ks8695_rx(ksp, budget);
> 
> 	if (work_done < budget) {
> 		unsigned long flags;
> 		spin_lock_irqsave(&ksp->rx_lock, flags);
> >>> packet arrives >>
> 		/*enable rx interrupt*/
> 		writel(isr | mask_bit, KS8695_IRQ_VA + KS8695_INTEN);
> >>> or here >>
> 		__napi_complete(napi);
> 		spin_unlock_irqrestore(&ksp->rx_lock, flags);
> 

yes, it is right, it maybe a race condition. so you can motify
ks8695_poll() to try, like this:

		__napi_complete(napi);
		/*enable rx interrupt*/
		writel(isr | mask_bit, KS8695_IRQ_VA + KS8695_INTEN); 


Best,
Figo.zhang

Re: KS8695: possible NAPI issue

[figo zhang]

2010/3/5 Yegor Yefremov <yegorslists@googlemail.com>:
>>> My tests look like following:
>>>
>>> 1. system start
>>> 2. ping one host: O.K.
>>> 3. nc -l -p 5000 > /var/test (about 1Mb): O.K.
>>> 4. ping the same host: failed
>> 1. type "arp" command, see the arp is exit or expire?
>
> debian:~# nc -l -p 5000 > /var/zImage
>
> debian:~# ping -c 1 192.168.1.38
>
> PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.
>
> From 192.168.1.66 icmp_seq=1 Destination Host Unreachable
>
>
>
> --- 192.168.1.38 ping statistics ---
>
> 1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
>
>
>
> debian:~# arp
>
> Address                  HWtype  HWaddress           Flags Mask            Iface
> 192.168.1.38          (incomplete)                              eth0
>
> 192.168.1.36             ether   00:10:18:39:19:aa   C                     eth0
>
>
>> 2. at this point, see is it still have RX interrpt and receive packet in
>> ks8695_rx()? (in some watchpoint add printk).
>
> I've inserted some printks and I can see that interrupts are coming
> and the received count will be increased. I can also see my system
> sending arp requests. Even this ping request is visible in wireshark
> and its reply, but the ping utility sees nothing of it.
>

when you netcat finished, it cannot ping, right? at this point, would
you like to add some printk at RX and TX?
i want to see the target board have send arp packet, or have receive
arp reply packet.

you can add such funtion to print the packet buffer:

void print_mem(unsigned char *p,u32 len, u8 * s)
{
	u32 i;
	printk("  %s   ram addr = %x , data len = %x",s, p, len);

	for (i=0;i<len;i++) {
		if (0==i%16) {
			printk("\n  0x%02x : ",(p+i));
		}
		printk("%02x",*(u8 *)(p+i));
		printk(" ");
	}
	printk("\n\n");
}

for TX: ks8695_start_xmit(struct sk_buff *skb, struct net_device *ndev)

=>
    	ksp->tx_ring[buff_n].status =
		cpu_to_le32(TDES_IC | TDES_FS | TDES_LS |
			    (skb->len & TDES_TBS));

	print_mem(skb->data, skb->len, "tx");

	wmb();

for RX: static int ks8695_rx(struct ks8695_priv *ksp, int budget)
=>
  			/* Retrieve the sk_buff */
			skb = ksp->rx_buffers[buff_n].skb;

			print_mem(skb->data, skb->len, "rx");

			/* Clear it from the ring */
			ksp->rx_buffers[buff_n].skb = NULL;

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

On Mon, Mar 8, 2010 at 10:04 AM, figo zhang <figo1802@gmail.com> wrote:
> 2010/3/5 Yegor Yefremov <yegorslists@googlemail.com>:
>>>> My tests look like following:
>>>>
>>>> 1. system start
>>>> 2. ping one host: O.K.
>>>> 3. nc -l -p 5000 > /var/test (about 1Mb): O.K.
>>>> 4. ping the same host: failed
>>> 1. type "arp" command, see the arp is exit or expire?
>>
>> debian:~# nc -l -p 5000 > /var/zImage
>>
>> debian:~# ping -c 1 192.168.1.38
>>
>> PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.
>>
>> From 192.168.1.66 icmp_seq=1 Destination Host Unreachable
>>
>>
>>
>> --- 192.168.1.38 ping statistics ---
>>
>> 1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
>>
>>
>>
>> debian:~# arp
>>
>> Address                  HWtype  HWaddress           Flags Mask            Iface
>> 192.168.1.38          (incomplete)                              eth0
>>
>> 192.168.1.36             ether   00:10:18:39:19:aa   C                     eth0
>>
>>
>>> 2. at this point, see is it still have RX interrpt and receive packet in
>>> ks8695_rx()? (in some watchpoint add printk).
>>
>> I've inserted some printks and I can see that interrupts are coming
>> and the received count will be increased. I can also see my system
>> sending arp requests. Even this ping request is visible in wireshark
>> and its reply, but the ping utility sees nothing of it.
>>
>
> when you netcat finished, it cannot ping, right? at this point, would
> you like to add some printk at RX and TX?
> i want to see the target board have send arp packet, or have receive
> arp reply packet.
>
> you can add such funtion to print the packet buffer:
>
> void print_mem(unsigned char *p,u32 len, u8 * s)
> {
>        u32 i;
>        printk("  %s   ram addr = %x , data len = %x",s, p, len);
>
>        for (i=0;i<len;i++) {
>                if (0==i%16) {
>                        printk("\n  0x%02x : ",(p+i));
>                }
>                printk("%02x",*(u8 *)(p+i));
>                printk(" ");
>        }
>        printk("\n\n");
> }
>
> for TX: ks8695_start_xmit(struct sk_buff *skb, struct net_device *ndev)
>
> =>
>        ksp->tx_ring[buff_n].status =
>                cpu_to_le32(TDES_IC | TDES_FS | TDES_LS |
>                            (skb->len & TDES_TBS));
>
>        print_mem(skb->data, skb->len, "tx");
>
>        wmb();
>
> for RX: static int ks8695_rx(struct ks8695_priv *ksp, int budget)
> =>
>                        /* Retrieve the sk_buff */
>                        skb = ksp->rx_buffers[buff_n].skb;
>
>                        print_mem(skb->data, skb->len, "rx");
>
>                        /* Clear it from the ring */
>                        ksp->rx_buffers[buff_n].skb = NULL;
>

skb->len in RX is always 0, so I took pktlen.

This is the end of the netcat transmission (the transfer stocked at
about 10% of file size):

  rx   ram addr = c288d020 , data len = 5ea

  dst:00:04:d9:80:94:cd src:00:18:f3:fe:84:84 type: 0x0800

  0xc288d02c : 45 00

  0xc288d030 : 05 dc 1a 4e 40 00 80 06 57 15 c0 a8 01 26 c0 a8

  0xc288d040 : 01 42 07 7d 13 88 db 12 9b 31 fe 7e 93 ae 50 10

  0xc288d050 : ff ff 60 69 00 00 d2 7c ed 11 b2 c9 3b 9a 6f 72

  0xc288d060 : 33 8d 3d f8 aa 97 30 3a 4b bc 19 93 39 e9 a2 e3

  0xc288d070 : d7 9a 28 d2 7e 48 f6 be a4 79 3a c3 31 13 2c 4d

  0xc288d080 : 0f 1b 6f 39 5d aa 6f c2 cd 81 29 88 2a 86 d2 56

  0xc288d090 : d0 7c 19 33 c1 5e 45 20 b9 ba dd 2a 7b f3 65 18

  0xc288d0a0 : 86 61 62 a4 a3 f9 de a8 93 8f 37 78 ec 87 b5 19

  0xc288d0b0 : 90 66 a2 a8 53 9c fd f7 75 67 89 34 85 1e 33 de

  0xc288d0c0 : 7e 10 b3 c4 7b 4c 62 cc 04 13 c5 fe 14 dc 44 7b

  0xc288d0d0 : 23 cd 62 0a d4 32 51 c5 50 da 0a 9a 2f 63 26 d8

  0xc288d0e0 : ab 08 24 57 b7 5b 65 6f be 81 ef 02 b9 98 0c 6e

  0xc288d0f0 : 60 06 4b 30 85 bb 9f ff dc fc c3 6c 6e 7d 7b 6e

  0xc288d100 : 43 e0 26 47 7b 20 7a 06 f1 29 ce bb 71 8a dd ee

  0xc288d110 : c6 1e f9 34 7e 1a 27 4d 7f bb d2 cf 72 dd 19 d3

  0xc288d120 : f8 4e 37 ed a4 f1 96 13 f5 31 a5 7f 3a c9 96 13

  0xc288d130 : 69 8a 2e d5 1e b0 06 8a a0 a8 62 28 6d 05 cd 97

  0xc288d140 : 31 13 ec 55 04 92 ab db ad 52 99 ef f1 78 6c 9a

  0xc288d150 : 66 b7 db d5 75 fd f2 db 1f 2e 9f 4f cf ee a0 e7

  0xc288d160 : 5d 20 17 93 c1 0d cc 60 09 a6 68 69 be f6 08 dc

  0xc288d170 : e4 68 0f 44 cf 20 3e c5 79 77 4d 09 24 2b e5 0d

  0xc288d180 : 8c 3d d7 6d 9d 34 cd 2c f1 dc ea ea 83 9f 1d 79

  0xc288d190 : a5 42 9a 66 22 cf a3 89 ba 07 ae 3e be 7f cc bc

  0xc288d1a0 : 3f e0 dd 57 11 14 55 0c a5 ad a0 f9 32 66 82 bd

  0xc288d1b0 : 8a 40 72 75 bb 55 7e 37 5f 67 b8 db ed b6 aa aa

  0xc288d1c0 : ba e7 a5 37 5f c4 fd 34 f5 2e 18 67 d3 73 eb ab

  0xc288d1d0 : 18 67 15 37 d0 4f 06 4b 30 45 4b f3 b5 47 e0 26

  0xc288d1e0 : 47 7b 20 7a 06 f1 29 36 fb 8d ba 0a b1 25 2b f1

  0xc288d1f0 : dd 8b 3d dd 6d 9d 34 cd 2c f1 ac 27 ea d3 19 1d

  0xc288d200 : 7c 8d 42 9a 66 22 bf 8d 5e 71 39 b7 d5 de 48 d3

  0xc288d210 : 99 02 45 50 54 31 94 b6 82 e6 cb 98 09 f6 2a 02

  0xc288d220 : c9 d5 ed 56 39 30 5f 68 af 13 de fd 7e 7f e8 79

  0xc288d230 : 7f 78 44 9c f9 ba b8 13 46 df 85 13 9b 01 17 bc

  0xc288d240 : 59 97 fd b9 ff 4f 42 3e 47 62 58 cf bf 2f 5e 75

  0xc288d250 : 9b 4a 1b f5 15 9e bd 6f 33 cf 4f b6 84 3b a1 a5

  0xc288d260 : f9 da 23 70 93 a3 3d 10 3d 83 f8 14 67 df 38 85

  0xc288d270 : 94 86 f9 83 c7 9b 42 4c 97 60 8a 04 6b 89 3d cb

  0xc288d280 : d2 ba 73 95 59 e2 6d 86 a9 31 63 cf 12 e9 31 25

  0xc288d290 : 98 68 f4 59 44 7d 34 a7 91 63 6c 03 cf 98 57 7f

  0xc288d2a0 : 3a d6 40 11 14 55 0c a5 ad a0 f9 32 66 82 bd 8a

  0xc288d2b0 : 40 72 75 bb 55 0e cc b7 69 9a aa aa 9c f6 be be

  0xc288d2c0 : be be f5 fc fa f1 13 79 7f 78 74 e6 fb f9 f4 3c

  0xc288d2d0 : fa 2e 80 f0 17 7f 33 a1 96 8b fe 70 fe 4f 73 86

  0xc288d2e0 : 9d ba a4 e1 37 fc e2 f1 bb 54 da a8 af dc 7f ab

  0xc288d2f0 : 17 9d 9f 66 09 f7 43 4b f3 b5 47 e0 26 47 7b 20

  0xc288d300 : 7a 06 f1 29 ce be 71 22 84 10 b2 52 50 04 45 15

  0xc288d310 : 43 69 2b 68 be 8c 99 60 af 22 90 5c dd 6e 95 03

  0xc288d320 : f3 dd ed 76 75 5d 1f 0e 07 e7 bc 1f 3d 4e 75 5d

  0xc288d330 : 4e e6 eb 32 fa 2e 74 d3 02 b8 f9 62 ea c8 50 ac

  0xc288d340 : a6 d4 52 9f af cf 11 27 7b fe 7c 8e cf 4e 0d 2e

  0xc288d350 : 7e 0d 6f 56 db 24 da b8 c8 64 97 9e 9f 66 09 4b

  0xc288d360 : 19 3e 68 ff 0e b4 46 4b f3 b5 47 e0 26 47 7b 20

  0xc288d370 : 7a 06 f1 29 36 bb 21 09 21 84 dc 39 28 82 a2 8a

  0xc288d380 : a1 b4 15 34 5f c6 4c b0 57 11 48 ae 6e b7 ca 81

  0xc288d390 : f9 d6 3d 27 f3 3d 1e 8f 7f 57 7f 85 98 ef 1c bd

  0xc288d3a0 : 1a d5 d2 39 9e eb b7 54 ff 69 fe 3f 99 3f 66 78

  0xc288d3b0 : b3 da 26 d1 c6 45 26 bb f4 fc 34 4b 58 8a 7e 46

  0xc288d3c0 : 7a 21 36 69 69 be f6 08 dc e4 68 0f 44 cf 20 3e

  0xc288d3d0 : c5 f6 77 26 21 84 90 fb 04 45 50 54 31 94 b6 82

  0xc288d3e0 : e6 cb 98 09 f6 2a 02 c9 d5 ed 56 39 30 5f b8 ad

  0xc288d3f0 : 93 5c a7 ba 4e 78 9d f6 1e 8f c7 8f 8f 8f b7 b7

  0xc288d400 : b7 c3 e1 50 d7 f5 6e b7 1b 7d 17 ba ef 7e 71 52

  0xc288d410 : d7 cd 77 c4 09 73 cc f4 ac 96 ea d9 2f 1b 41 33

  0xc288d420 : 67 cc d1 55 2f a2 4d a5 8d 67 1f 84 ff 7c cf ad

  0xc288d430 : 4b b6 84 45 8c 3e a3 ee 6b 5d 37 bb ac 19 b4 34

  0xc288d440 : 5f 7b 04 6e 72 b4 07 a2 67 10 9f 62 e3 db 92 10

  0xc288d450 : 42 c8 dd 82 22 28 aa 18 4a 5b 41 f3 65 cc 04 7b

  0xc288d460 : 15 81 e4 ea 76 ab 9c 36 5f 97 8f 9e 45 e6 3b fc

  0xc288d470 : af ee e5 96 9a e9 4c 6f 9d 63 a9 fe 11 f4 75 0a

  0xc288d480 : 46 c7 9c ba 8c f9 b4 c9 b5 d1 ff 50 a6 ce f7 dc

  0xc288d490 : ba f4 4b 98 83 7e 46 62 09 66 45 a3 a5 f9 da 23

  0xc288d4a0 : 70 93 a3 3d 10 3d 83 f8 14 9b dd 90 84 10 42 3c

  0xc288d4b0 : 88 46 71 e6 4f eb 02 45 50 54 31 94 b6 82 e6 cb

  0xc288d4c0 : 98 09 f6 2a 02 c9 d5 ed 56 39 30 5f a7 ba 30 df

  0xc288d4d0 : 7f ff fe e7 af 1f 3f 5d de 7a 5e 5f 5f f7 fb 7d

  0xc288d4e0 : 55 55 4d d3 8c be 0b 60 8e 75 2e 35 d3 d1 5f cf

  0xc288d4f0 : fe 34 7f fc 39 07 3d 83 f8 7f 9a 43 9b 56 1b fd

  0xc288d500 : da eb f9 d5 73 5a e2 25 cc 44 5f b9 e7 c1 99 a2

  0xc288d510 : a5 f9 da 23 70 93 a3 3d 10 3d 83 f8 14 9b dd 90

  0xc288d520 : 84 10 42 3c 9c fd 7a 67 f0 79 47 11 14 55 0c a5

  0xc288d530 : ad a0 f9 32 66 82 bd 8a 40 72 75 bb 55 0e cc f7

  0xc288d540 : 78 3c 3a f3 75 79 7f 78 44 0e 3d 4e 7b eb ba de

  0xc288d550 : 6e b7 ee 84 d1 77 e1 c4 66 80 38 32 75 8e e7 a7

  0xc288d560 : d1 11 46 dd 76 6a b4 6e cc 77 f4 f8 fa df 8b f4

  0xc288d570 : 76 f4 b2 67 d2 a6 d2 46 7d 79 a3 47 ce 9e af d7

  0xc288d580 : 98 6c 09 8b d0 8f 52 ef 16 9b b4 34 5f 7b 04 6e

  0xc288d590 : 72 b4 07 a2 67 10 9f 62 cb 7b 92 10 42 c8 14 67

  0xc288d5a0 : bf de 19 7c de 51 04 45 15 43 69 2b 68 be 8c 99

  0xc288d5b0 : 60 af 22 90 5c dd 6e 95 03 f3 75 38 b7 75 e6 7b

  0xc288d5c0 : 4a dd 53 55 d5 a8 f6 76 56 ad 67 45 64 70 03 33

  0xc288d5d0 : 58 82 29 5a 9a af 3d 02 37 39 da 03 d1 33 88 4f

  0xc288d5e0 : 71 06 ad 11 21 84 dc 21 e2 eb bd f9 62 ea 84 35

  0xc288d5f0 : 82 22 28 aa 18 4a 5b 41 f3 65 cc 04 7b 15 81 e4

  0xc288d600 : ea 76 ab fc 6e be 0e 67 b8 9f



  rx   ram addr = c2853020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x0806

  0xc285302c : 00 01

  0xc2853030 : 08 00 06 04 00 01 00 18 f3 fe 84 84 c0 a8 01 26

  0xc2853040 : 00 00 00 00 00 00 c0 a8 01 03 00 00 00 00 00 00

  0xc2853050 : 00 00 00 00 00 00 00 00 00 00 00 00



  rx   ram addr = c2846020 , data len = 5ea

  dst:00:04:d9:80:94:cd src:00:18:f3:fe:84:84 type: 0x0800

  0xc284602c : 45 00

  0xc2846030 : 05 dc 1a 51 40 00 80 06 57 12 c0 a8 01 26 c0 a8

  0xc2846040 : 01 42 07 7d 13 88 db 12 b7 b5 fe 7e 93 ae 50 10

  0xc2846050 : ff ff b6 b5 00 00 6d 2a f5 af 3a 5b 3e 26 ff 05

  0xc2846060 : c3 6a 7e 9d 42 c6 aa 25 05 46 01 8b c1 c9 c0 ce

  0xc2846070 : 52 f9 12 42 08 e9 4d 9c fe b0 df 46 d3 d2 20 4d

  0xc2846080 : d1 5b 91 1b 70 64 56 12 a9 54 cc a9 ea 0e 75 79

  0xc2846090 : 03 81 30 87 5e 38 24 4e 65 53 cd 93 82 34 bf a5

  0xc28460a0 : 9e cb 19 c0 df 24 44 ab fd 71 b9 cc e4 ad c3 55

  0xc28460b0 : e7 40 c6 4a 2e 72 63 70 06 02 8b c1 c9 c0 ce 52

  0xc28460c0 : f9 12 42 08 e9 4d 9c fe b0 df 46 d3 d2 20 4d d1

  0xc28460d0 : 5b 91 1b 70 64 56 12 a9 54 cc 69 54 31 49 13 08

  0xc28460e0 : 52 44 cc 4d 9c ca 86 ed 94 fa d4 28 ee 6a d2 b5

  0xc28460f0 : f8 d0 72 98 ba 0c 8c 06 14 6f c1 26 9d 45 c6 4a

  0xc2846100 : 5d c6 57 bd 2e 08 aa d7 b3 e6 29 4e 07 2a 5f 42

  0xc2846110 : 08 21 bd 89 d3 1f f6 db 68 5a 5a 88 19 bb 0f f2

  0xc2846120 : 0c 1c 99 95 44 2a 15 73 d6 56 32 1e 09 34 ce c4

  0xc2846130 : c4 a9 6c d8 4e 2c e8 c0 c7 52 3f ab 5d 2a 82 55

  0xc2846140 : 61 5f 3c 85 40 6b 5c 5a aa 06 2c ce 2f 7c 5d 10

  0xc2846150 : 0e 57 c8 4c 60 67 a9 7c 09 21 84 f4 26 4e 7f d8

  0xc2846160 : 6f a3 69 69 90 a6 e8 ad c8 0d 38 32 2b 89 54 2a

  0xc2846170 : e6 3c 94 27 f2 24 1d 89 97 59 89 53 d9 b0 9d 46

  0xc2846180 : 7d 6a 5c 0c 76 a9 08 56 c5 a9 c5 73 a1 14 95 b1

  0xc2846190 : b2 bb 33 2e 87 2b 64 26 b0 b3 54 be 84 10 42 7a

  0xc28461a0 : 13 a7 3f ec b7 d1 b4 34 48 53 f4 56 e4 06 1c 99

  0xc28461b0 : 95 44 2a e5 13 6e 6f e0 2b e0 ae 1c 33 31 71 2a

  0xc28461c0 : 9b b4 6d cb 28 2e d6 06 a8 df 51 fd c4 ea b3 60

  0xc28461d0 : 66 39 bf f1 1c 4f e2 00 c7 0a bf 7a 5c b0 18 9c

  0xc28461e0 : 0c ec ac a6 7c ff b3 1f 3f 76 e5 fb f5 cf fd f8

  0xc28461f0 : f2 f2 af 4f bf 7d a3 f2 25 84 10 e2 26 4e 7f d8

  0xc2846200 : 6f a3 69 e9 91 66 69 af c8 b3 71 64 56 12 a9 d4

  0xc2846210 : c3 b0 69 24 03 26 4e 65 bb 37 0e 46 2c ab e2 70

  0xc2846220 : 4c e3 d2 1a 25 56 d7 b2 94 d7 d8 59 2a 5f 42 08

  0xc2846230 : 21 bd 89 d3 1f f6 db 68 5a fa b1 35 54 02 b9 1e

  0xc2846240 : 47 66 25 91 4a 3d 0c eb 34 6d 34 e2 54 b6 7b e3

  0xc2846250 : 60 64 7b 03 8f 69 79 fc 90 51 62 75 2d 4b 79 8d

  0xc2846260 : 9d 9d 58 f9 6e 19 c5 45 e3 b0 45 ea f6 93 b1 47

  0xc2846270 : 95 f1 27 64 1a e2 f4 87 d7 6e 34 39 2d 25 8b d5

  0xc2846280 : 8e f8 70 64 56 12 a9 d4 c3 b0 45 76 f0 38 95 8d

  0xc2846290 : 07 0f 1e 96 63 38 e5 bb d7 84 77 7f 83 8b 6a 21

  0xc28462a0 : 63 6b b7 8a 2b e0 29 d2 0f 86 9a 90 e9 89 d3 1f

  0xc28462b0 : 5a 36 1a 1f 2d a5 8c 55 90 f8 70 64 56 12 a9 d4

  0xc28462c0 : c3 b0 45 76 f6 38 95 ed de 38 0c c4 9a b1 5a ca

  0xc28462d0 : 6b ec ec 70 ca f7 55 f3 ee ec 27 60 58 ad e4 be

  0xc28462e0 : 5e 2f ee 3e fe e6 d7 17 29 da 76 5e c3 9e 9f 14

  0xc28462f0 : b7 6a 17 f3 f0 e2 61 78 a4 bc 0e 4c 22 84 84 22

  0xc2846300 : 4e 7f 68 d9 68 7c b4 54 21 16 30 e2 c3 91 59 49

  0xc2846310 : a4 52 0f c3 16 d9 94 e3 54 b6 7b e3 30 10 6b c6

  0xc2846320 : 6a 29 af b1 b3 54 be f9 df fc fa 22 45 fb 14 85

  0xc2846330 : ea 2c 4e 0e 2f aa 7f d5 2b 87 0f e6 2f 52 5f 4a

  0xc2846340 : 08 09 45 9c fe d0 b2 d1 f8 68 29 41 ac 5e c4 87

  0xc2846350 : 23 b3 92 48 a5 1e 86 2d b2 23 c7 a9 6c 2d 5e 3c

  0xc2846360 : ed 63 6d 6f b4 cf e3 9e 6d 29 0d f8 60 29 af b1

  0xc2846370 : b3 c3 29 df 57 cd bb 83 97 3a 16 53 52 52 c9 dc

  0xc2846380 : 59 a4 68 9f e2 94 c8 ad 55 24 cb 15 f5 fa a9 b7

  0xc2846390 : 13 42 42 11 a7 3f b4 6c 34 3e 5a 4a 10 ab 17 f1

  0xc28463a0 : e1 c8 ac 24 52 a9 87 61 8b ec c8 71 2a 9b 6a de

  0xc28463b0 : 96 91 cc ed 16 98 01 0c b3 84 cb de b3 d9 ef 9e

  0xc28463c0 : 5d 69 20 56 87 17 8b 38 58 22 13 04 2c 06 27 03

  0xc28463d0 : 3b 3b 9c f2 35 52 5b c0 6a 0a e3 d5 3e c4 92 7e

  0xc28463e0 : 0e 3e ed 09 04 2c 78 0a 3f 68 7f 3b 21 24 02 71

  0xc28463f0 : fa c3 0b 37 9a 82 96 12 c4 ea 45 7c 38 32 2b 89

  0xc2846400 : 54 ea 61 d8 22 3b 72 9c ca 26 6d 73 b7 5b b5 5b

  0xc2846410 : c6 61 00 7b cf 66 bf 7b 76 a5 d5 62 a5 86 0b bc

  0xc2846420 : 68 ac fe 13 8b c1 c9 c0 ce ae a0 7c 0f 55 12 d6

  0xc2846430 : 6b f1 d7 f3 d3 b0 47 d5 58 2b f0 60 4b 91 19 ab

  0xc2846440 : f2 10 b2 2c 71 fa c3 0b 37 9a 82 96 12 c4 ea 45

  0xc2846450 : 7c 38 32 2b 89 54 ea 61 d8 22 3b 72 9c ca 26 6d

  0xc2846460 : ab 49 b9 bc 83 2a ae 1c 4e 22 9f b5 fc 55 67 3b

  0xc2846470 : 35 83 fc 0b dc c4 d4 64 91 ea 29 d0 02 63 f5 9f

  0xc2846480 : 58 0c 4e 06 76 56 2a df 9f 3f 7f ed 47 a6 7c bf

  0xc2846490 : 7f fa fc 6d 38 e5 9b 60 d6 d4 ce e5 b3 b8 20 ac

  0xc28464a0 : 46 2d 26 f2 a2 3a 12 44 b5 f6 b1 6a 33 c8 bf fc

  0xc28464b0 : 58 84 44 26 4e 7f 78 ed 46 93 d3 52 7f 58 ba 88

  0xc28464c0 : 0f 47 66 25 91 4a 3d 0c 5b 64 3b 8e 53 d9 54 f3

  0xc28464d0 : 64 d7 94 df 52 cf e5 0c e0 6f 12 0a d1 fe 38 3e

  0xc28464e0 : c7 13 1a 8d 57 01 b1 92 e7 6a 5f 8a ed 8f 09 16

  0xc28464f0 : 83 93 81 9d 9d 58 f9 92 cb 89 9f da 84 90 98 c4

  0xc2846500 : e9 0f fb 6d 34 2d 15 92 c5 95 f8 70 64 56 12 a9

  0xc2846510 : d4 c3 b0 45 1a 86 38 95 0d db 89 05 23 f8 58 35

  0xc2846520 : e9 5a 28 41 39 ac d0 89 f8 a5 f2 1c 4f d8 a2 3a

  0xc2846530 : 6b b1 2a e6 3c 7c 63 8b 0d cf 47 7a ad 7e a0 39

  0xc2846540 : c0 e9 40 e5 4b 8c 4c 9c 23 84 90 de c4 e9 0f d5

  0xc2846550 : 8d a6 d6 65 a9 be d4 ee b6 94 47 56 56 e2 c3 91

  0xc2846560 : 59 49 a4 52 0f c3 16 e9 16 e2 54 36 6c 27 16 6b

  0xc2846570 : e0 63 a9 d2 d5 2e 03 d5 c7 2d e7 be bb 16 6a b1

  0xc2846580 : 52 e7 79 5c bc d6 86 e7 73 b8 42 66 02 3b 4b e5

  0xc2846590 : 4b 08 21 a4 37 71 fa 43 b9 d1 14 32 16 f7 33 c6

  0xc28465a0 : a6 ee 2c e1 fb 26 12 14 47 66 25 91 4a 3d 0c 8b

  0xc28465b0 : af 05 2e 21 4e 65 c3 76 1a f5 a9 b1 e2 9d 15 aa

  0xc28465c0 : 87 b7 ce 4e d8 a2 3a 6b b1 52 e7 71 b8 1c 93 c3

  0xc28465d0 : 15 32 13 d8 59 2a 5f 42 08 21 bd 89 d3 1f aa 1b

  0xc28465e0 : 8d b1 6d c3 77 5b 9a 9f f0 7d 13 09 8a 23 b3 92

  0xc28465f0 : 48 a5 1e 86 c5 d7 02 97 10 a7 b2 49 db b6 8c e2

  0xc2846600 : 62 6d 80 5a f1 e4 24 b5 67 c1



  tx   ram addr = c36c389e , data len = 42

  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800

  0xc36c38aa : 45 00

  0xc36c38ae : 00 34 39 c5 40 00 40 06 7d 46 c0 a8 01 42 c0 a8

  0xc36c38be : 01 26 13 88 07 7d fe 7e 93 ae db 12 b2 01 80 10

  0xc36c38ce : 0b 68 85 10 00 00 01 01 05 0a db 12 b7 b5 db 12

  0xc36c38de : bd 69



  rx   ram addr = c283d020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x0806

  0xc283d02c : 00 01

  0xc283d030 : 08 00 06 04 00 01 00 18 f3 fe 84 84 c0 a8 01 26

  0xc283d040 : 00 00 00 00 00 00 c0 a8 01 01 00 00 00 00 00 00

  0xc283d050 : 00 00 00 00 00 00 00 00 00 00 00 00



  tx   ram addr = c35cd09e , data len = 42

  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800

  0xc35cd0aa : 45 00

  0xc35cd0ae : 00 34 39 c6 40 00 40 06 7d 45 c0 a8 01 42 c0 a8

  0xc35cd0be : 01 26 13 88 07 7d fe 7e 93 ae db 12 b2 01 80 10

  0xc35cd0ce : 22 38 6e 40 00 00 01 01 05 0a db 12 b7 b5 db 12

  0xc35cd0de : bd 69



  tx   ram addr = c35cd09e , data len = 42

  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800

  0xc35cd0aa : 45 00

  0xc35cd0ae : 00 34 39 c7 40 00 40 06 7d 44 c0 a8 01 42 c0 a8

  0xc35cd0be : 01 26 13 88 07 7d fe 7e 93 ae db 12 b2 01 80 10

  0xc35cd0ce : 44 70 4c 08 00 00 01 01 05 0a db 12 b7 b5 db 12

  0xc35cd0de : bd 69



  tx   ram addr = c35cd09e , data len = 42

  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800

  0xc35cd0aa : 45 00

  0xc35cd0ae : 00 34 39 c8 40 00 40 06 7d 43 c0 a8 01 42 c0 a8

  0xc35cd0be : 01 26 13 88 07 7d fe 7e 93 ae db 12 b2 01 80 10

  0xc35cd0ce : 88 e0 07 98 00 00 01 01 05 0a db 12 b7 b5 db 12

  0xc35cd0de : bd 69



  rx   ram addr = c2847020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x0806

  0xc284702c : 00 01

  0xc2847030 : 08 00 06 04 00 01 00 18 f3 fe 84 84 c0 a8 01 26

  0xc2847040 : 00 00 00 00 00 00 c0 a8 01 01 00 00 00 00 00 00

  0xc2847050 : 00 00 00 00 00 00 00 00 00 00 00 00



  rx   ram addr = c287f020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x0806

  0xc287f02c : 00 01

  0xc287f030 : 08 00 06 04 00 01 00 18 f3 fe 84 84 c0 a8 01 26

  0xc287f040 : 00 00 00 00 00 00 c0 a8 01 03 00 00 00 00 00 00

  0xc287f050 : 00 00 00 00 00 00 00 00 00 00 00 00



  rx   ram addr = c2892020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x002c

  0xc289202c : e0 e0

  0xc2892030 : 03 ff ff 00 28 00 01 00 00 00 00 ff ff ff ff ff

  0xc2892040 : ff 04 53 00 00 00 00 00 18 f3 fe 84 84 04 53 00

  0xc2892050 : 02 c1 42 fd 3b 00 01 00 02 00 00 00



  rx   ram addr = c2891020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x0806

  0xc289102c : 00 01

  0xc2891030 : 08 00 06 04 00 01 00 18 f3 fe 84 84 c0 a8 01 26

  0xc2891040 : 00 00 00 00 00 00 c0 a8 01 01 00 00 00 00 00 00

  0xc2891050 : 00 00 00 00 00 00 00 00 00 00 00 00



  rx   ram addr = c2890020 , data len = 3c

  dst:ff:ff:ff:ff:ff:ff src:00:18:f3:fe:84:84 type: 0x0806

  0xc289002c : 00 01

  0xc2890030 : 08 00 06 04 00 01 00 18 f3 fe 84 84 c0 a8 01 26

  0xc2890040 : 00 00 00 00 00 00 c0 a8 01 03 00 00 00 00 00 00

  0xc2890050 : 00 00 00 00 00 00 00 00 00 00 00 00

Yegor

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

This is ping afterwards:

debian:~# ping -c 1 192.168.1.38

PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.


  tx   ram addr = c371c202 , data len = 62

  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800

  0xc371c20e : 45 00

  0xc371c212 : 00 54 00 00 40 00 40 01 b6 f0 c0 a8 01 42 c0 a8

  0xc371c222 : 01 26 08 00 65 51 1a 07 00 01 92 00 95 4b 5f 57

  0xc371c232 : 07 00 08 09 0a 0b 0c 0d 0e 0f 10 11 12 13 14 15

  0xc371c242 : 16 17 18 19 1a 1b 1c 1d 1e 1f 20 21 22 23 24 25

  0xc371c252 : 26 27 28 29 2a 2b 2c 2d 2e 2f 30 31 32 33 34 35

  0xc371c262 : 36 37



  tx   ram addr = c371c202 , data len = 2a

  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0806

  0xc371c20e : 00 01

  0xc371c212 : 08 00 06 04 00 01 00 04 d9 80 94 cd c0 a8 01 42

  0xc371c222 : 00 00 00 00 00 00 c0 a8 01 26



  rx   ram addr = c35d2020 , data len = 3c

  dst:00:04:d9:80:94:cd src:00:18:f3:fe:84:84 type: 0x0806

  0xc35d202c : 00 01

  0xc35d2030 : 08 00 06 04 00 02 00 18 f3 fe 84 84 c0 a8 01 26

  0xc35d2040 : 00 04 d9 80 94 cd c0 a8 01 42 00 00 00 00 00 00

  0xc35d2050 : 00 00 00 00 00 00 00 00 00 00 00 00





--- 192.168.1.38 ping statistics ---

1 packets transmitted, 0 received, 100% packet loss, time 0ms

Yegor

Re: KS8695: possible NAPI issue

[figo zhang]

2010/3/8 Yegor Yefremov <yegorslists@googlemail.com>:
> This is ping afterwards:
>
> debian:~# ping -c 1 192.168.1.38
>
> PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.
>
>
>  tx   ram addr = c371c202 , data len = 62
>
>  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800

=> the print info is not clear, it had better just keep "print_mem()",
and remove other printk.

>
>  0xc371c20e : 45 00
>
>  0xc371c212 : 00 54 00 00 40 00 40 01 b6 f0 c0 a8 01 42 c0 a8
>
>  0xc371c222 : 01 26 08 00 65 51 1a 07 00 01 92 00 95 4b 5f 57
>
>  0xc371c232 : 07 00 08 09 0a 0b 0c 0d 0e 0f 10 11 12 13 14 15
>
>  0xc371c242 : 16 17 18 19 1a 1b 1c 1d 1e 1f 20 21 22 23 24 25
>
>  0xc371c252 : 26 27 28 29 2a 2b 2c 2d 2e 2f 30 31 32 33 34 35
>
>  0xc371c262 : 36 37
>
>
>
>  tx   ram addr = c371c202 , data len = 2a
>
>  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0806
>
>  0xc371c20e : 00 01
>
>  0xc371c212 : 08 00 06 04 00 01 00 04 d9 80 94 cd c0 a8 01 42
=> it is a send arp packet, "c0 a8 01 42" is your curr ip addr
192.168.1.66, tell the other side it's ip addr.
>
>  0xc371c222 : 00 00 00 00 00 00 c0 a8 01 26
>
>
>
>  rx   ram addr = c35d2020 , data len = 3c
>
>  dst:00:04:d9:80:94:cd src:00:18:f3:fe:84:84 type: 0x0806
>
>  0xc35d202c : 00 01
>
>  0xc35d2030 : 08 00 06 04 00 02 00 18 f3 fe 84 84 c0 a8 01 26

=> it is a reply arp packet, "c0 a8 01 26" (192.168.1.38 ) is other
side ip addr. so the arp
request is finished.

>
>  0xc35d2040 : 00 04 d9 80 94 cd c0 a8 01 42 00 00 00 00 00 00
>
>  0xc35d2050 : 00 00 00 00 00 00 00 00 00 00 00 00
>
=> so , it should still send  "IMCP" packet, you using "ping -c 1", it
send one IMCP packet , and wait for reply.  at this point, it have not
send packet?

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

On Tue, Mar 9, 2010 at 2:50 AM, figo zhang <figo1802@gmail.com> wrote:
> 2010/3/8 Yegor Yefremov <yegorslists@googlemail.com>:
>> This is ping afterwards:
>>
>> debian:~# ping -c 1 192.168.1.38
>>
>> PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.
>>
>>
>>  tx   ram addr = c371c202 , data len = 62
>>
>>  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800
>
> => the print info is not clear, it had better just keep "print_mem()",
> and remove other printk.
>
>>
>>  0xc371c20e : 45 00
>>
>>  0xc371c212 : 00 54 00 00 40 00 40 01 b6 f0 c0 a8 01 42 c0 a8
>>
>>  0xc371c222 : 01 26 08 00 65 51 1a 07 00 01 92 00 95 4b 5f 57
>>
>>  0xc371c232 : 07 00 08 09 0a 0b 0c 0d 0e 0f 10 11 12 13 14 15
>>
>>  0xc371c242 : 16 17 18 19 1a 1b 1c 1d 1e 1f 20 21 22 23 24 25
>>
>>  0xc371c252 : 26 27 28 29 2a 2b 2c 2d 2e 2f 30 31 32 33 34 35
>>
>>  0xc371c262 : 36 37
>>
>>
>>
>>  tx   ram addr = c371c202 , data len = 2a
>>
>>  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0806
>>
>>  0xc371c20e : 00 01
>>
>>  0xc371c212 : 08 00 06 04 00 01 00 04 d9 80 94 cd c0 a8 01 42
> => it is a send arp packet, "c0 a8 01 42" is your curr ip addr
> 192.168.1.66, tell the other side it's ip addr.
>>
>>  0xc371c222 : 00 00 00 00 00 00 c0 a8 01 26
>>
>>
>>
>>  rx   ram addr = c35d2020 , data len = 3c
>>
>>  dst:00:04:d9:80:94:cd src:00:18:f3:fe:84:84 type: 0x0806
>>
>>  0xc35d202c : 00 01
>>
>>  0xc35d2030 : 08 00 06 04 00 02 00 18 f3 fe 84 84 c0 a8 01 26
>
> => it is a reply arp packet, "c0 a8 01 26" (192.168.1.38 ) is other
> side ip addr. so the arp
> request is finished.
>
>>
>>  0xc35d2040 : 00 04 d9 80 94 cd c0 a8 01 42 00 00 00 00 00 00
>>
>>  0xc35d2050 : 00 00 00 00 00 00 00 00 00 00 00 00
>>
> => so , it should still send  "IMCP" packet, you using "ping -c 1", it
> send one IMCP packet , and wait for reply.  at this point, it have not
> send packet?

I made some tests on Fr with a Netbook connected directly to our
ks8695 based device and not to the companies network. During the test
I couldn't encounter any problem. Than I connected ks8695 again to the
companies network and the issue was there right away. So I decided to
check how many broadcast I get and it turned out that I have one
device on the network that is sending broadcast almost every second if
not more frequent. It turned out that the network on ks8695 was not
completely down, but very slow and many packets got lost. As soon as I
removed the broadcast sending device I could ping ks8695. The
consequences:

1. as long as I make no netcat transfer, the network responses are as
usual regardless of many broadcasts
2. if during netcat a broadcast occurs, than the network is getting
slow, so that if during a constant ping you get a broadcast that ping
response gets lost

Any idea how broadcast during transfer could affect NAPI in that way?

Regards,
Yeor

Re: KS8695: possible NAPI issue

[Yegor Yefremov]

On Mon, Mar 15, 2010 at 11:19 AM, Yegor Yefremov
<yegorslists@googlemail.com> wrote:
> On Tue, Mar 9, 2010 at 2:50 AM, figo zhang <figo1802@gmail.com> wrote:
>> 2010/3/8 Yegor Yefremov <yegorslists@googlemail.com>:
>>> This is ping afterwards:
>>>
>>> debian:~# ping -c 1 192.168.1.38
>>>
>>> PING 192.168.1.38 (192.168.1.38) 56(84) bytes of data.
>>>
>>>
>>>  tx   ram addr = c371c202 , data len = 62
>>>
>>>  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0800
>>
>> => the print info is not clear, it had better just keep "print_mem()",
>> and remove other printk.
>>
>>>
>>>  0xc371c20e : 45 00
>>>
>>>  0xc371c212 : 00 54 00 00 40 00 40 01 b6 f0 c0 a8 01 42 c0 a8
>>>
>>>  0xc371c222 : 01 26 08 00 65 51 1a 07 00 01 92 00 95 4b 5f 57
>>>
>>>  0xc371c232 : 07 00 08 09 0a 0b 0c 0d 0e 0f 10 11 12 13 14 15
>>>
>>>  0xc371c242 : 16 17 18 19 1a 1b 1c 1d 1e 1f 20 21 22 23 24 25
>>>
>>>  0xc371c252 : 26 27 28 29 2a 2b 2c 2d 2e 2f 30 31 32 33 34 35
>>>
>>>  0xc371c262 : 36 37
>>>
>>>
>>>
>>>  tx   ram addr = c371c202 , data len = 2a
>>>
>>>  dst:00:18:f3:fe:84:84 src:00:04:d9:80:94:cd type: 0x0806
>>>
>>>  0xc371c20e : 00 01
>>>
>>>  0xc371c212 : 08 00 06 04 00 01 00 04 d9 80 94 cd c0 a8 01 42
>> => it is a send arp packet, "c0 a8 01 42" is your curr ip addr
>> 192.168.1.66, tell the other side it's ip addr.
>>>
>>>  0xc371c222 : 00 00 00 00 00 00 c0 a8 01 26
>>>
>>>
>>>
>>>  rx   ram addr = c35d2020 , data len = 3c
>>>
>>>  dst:00:04:d9:80:94:cd src:00:18:f3:fe:84:84 type: 0x0806
>>>
>>>  0xc35d202c : 00 01
>>>
>>>  0xc35d2030 : 08 00 06 04 00 02 00 18 f3 fe 84 84 c0 a8 01 26
>>
>> => it is a reply arp packet, "c0 a8 01 26" (192.168.1.38 ) is other
>> side ip addr. so the arp
>> request is finished.
>>
>>>
>>>  0xc35d2040 : 00 04 d9 80 94 cd c0 a8 01 42 00 00 00 00 00 00
>>>
>>>  0xc35d2050 : 00 00 00 00 00 00 00 00 00 00 00 00
>>>
>> => so , it should still send  "IMCP" packet, you using "ping -c 1", it
>> send one IMCP packet , and wait for reply.  at this point, it have not
>> send packet?
>
> I made some tests on Fr with a Netbook connected directly to our
> ks8695 based device and not to the companies network. During the test
> I couldn't encounter any problem. Than I connected ks8695 again to the
> companies network and the issue was there right away. So I decided to
> check how many broadcast I get and it turned out that I have one
> device on the network that is sending broadcast almost every second if
> not more frequent. It turned out that the network on ks8695 was not
> completely down, but very slow and many packets got lost. As soon as I
> removed the broadcast sending device I could ping ks8695. The
> consequences:
>
> 1. as long as I make no netcat transfer, the network responses are as
> usual regardless of many broadcasts
> 2. if during netcat a broadcast occurs, than the network is getting
> slow, so that if during a constant ping you get a broadcast that ping
> response gets lost
>
> Any idea how broadcast during transfer could affect NAPI in that way?

Tried 2.6.35. Have still the same problems. Another observation:

if I add eth0 to a bridge (br0), then I have no problems. So ks8695
network driver is still involved, but the upper level seems to make
the difference.

Any idea?

Regards,
Yegor