[PLEASE REVERT] "Add a WARN_ON_ONCE" at net/sched/sch_generic.c:

[PLEASE REVERT] "Add a WARN_ON_ONCE" at net/sched/sch_generic.c:

[Komuro]

Hi,

Please revert the patch "Add a WARN_ON_ONCE() to
 the transmit timeout function".

The transmit timeout sometimes happens on normal operation.
This message annoys me.


>net: Add a WARN_ON_ONCE() to the transmit timeout function
>
>WARN_ON_ONCE() gives a stack trace including the full module list.
>Having this in the kernel dump for the timeout case in the
>generic netdev watchdog will help us see quicker which driver
>is involved. It also allows us to collect statistics
>and patterns in terms of which drivers have this event occuring.
>
>Suggested by Andrew Morton
>
>Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
>Signed-off-by: David S. Miller <davem@davemloft.net>

Best Regards
Komuro

Re: [PLEASE REVERT] "Add a WARN_ON_ONCE" at net/sched/sch_generic.c:

[David Miller]

From: Komuro <komurojun-mbn@nifty.com>
Date: Sun, 25 May 2008 11:23:08 +0900

> Please revert the patch "Add a WARN_ON_ONCE() to
>  the transmit timeout function".
> 
> The transmit timeout sometimes happens on normal operation.
> This message annoys me.

Sometimes, but not often.

Most of the time it does indicate a bug and we absolutely cannot
diagnose it without the backtrace.

Re: [PLEASE REVERT] "Add a WARN_ON_ONCE" at net/sched/sch_generic.c:

[Komuro]

On Sun, 25 May 2008 05:10:32 -0700 (PDT)
David Miller <davem@davemloft.net> wrote:

> 
> > Please revert the patch "Add a WARN_ON_ONCE() to
> >  the transmit timeout function".
> > 
> > The transmit timeout sometimes happens on normal operation.
> > This message annoys me.
> 
> Sometimes, but not often.
> 
> Most of the time it does indicate a bug and we absolutely cannot
> diagnose it without the backtrace.

I think the transmit timeout indicates 
that the cable is not connected or the irq is conflict.


Best Regards
Komuro

Re: [PLEASE REVERT] "Add a WARN_ON_ONCE" at net/sched/sch_generic.c:

[David Miller]

From: Komuro <komurojun-mbn@nifty.com>
Date: Tue, 27 May 2008 21:34:24 +0900

> I think the transmit timeout indicates 
> that the cable is not connected or the irq is conflict.

The MAC of the chip transmits and ACKs the packet in the hardware even
if the link is down.  The packet is simply dropped.

Also, when the link goes down, the driver should mark the device as
such.  In such a state, the generic networking does not pass new
packets to the driver.

Therefore, transmit timeouts in this case are a hardware
or driver error.

If there is an IRQ issue, that is an error that'd we would like to
have as much information as possible to diagnose.  So the warning is
still entirely appropriate.

I sense that you are seeing this warning triggered, and if so I
would suggest that you investigate it fully as it should not be
happening under normal circumstances.

RE: NETDEV WATCHDOG: wifi0: transmit timed out

[Komuro]

Hi,

Please try to comment out the "WARN_ON_ONCE"
at net/sched/sched_generic.c.


--- linux-2.6.27-rc1/net/sched/sch_generic.c.orig	2008-07-29 21:45:36.000000000 +0900
+++ linux-2.6.27-rc1/net/sched/sch_generic.c	2008-07-29 21:45:54.000000000 +0900
@@ -216,7 +216,6 @@ static void dev_watchdog(unsigned long a
 				printk(KERN_INFO "NETDEV WATCHDOG: %s (%s): transmit timed out\n",
 				       dev->name, netdev_drivername(dev, drivername, 64));
 				dev->tx_timeout(dev);
-				WARN_ON_ONCE(1);
 			}
 			if (!mod_timer(&dev->watchdog_timer,
 				       round_jiffies(jiffies +



>I should'nt be posting this, due to
>having  tainted modules(madwifi), but
>I figured I might as well:(below)
>
>[ 5000.986044] NETDEV WATCHDOG: wifi0: transmit timed out
>[ 5000.991959] ------------[ cut here ]------------
>[ 5000.991964] WARNING: at net/sched/sch_generic.c:222 dev_watchdog+0x95/0xe7()

Best Regards
Komuro

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Justin Mattock]

On Tue, Jul 29, 2008 at 5:53 AM, Komuro <komurojun-mbn@nifty.com> wrote:
> Hi,
>
> Please try to comment out the "WARN_ON_ONCE"
> at net/sched/sched_generic.c.
>
>
> --- linux-2.6.27-rc1/net/sched/sch_generic.c.orig       2008-07-29 21:45:36.000000000 +0900
> +++ linux-2.6.27-rc1/net/sched/sch_generic.c    2008-07-29 21:45:54.000000000 +0900
> @@ -216,7 +216,6 @@ static void dev_watchdog(unsigned long a
>                                printk(KERN_INFO "NETDEV WATCHDOG: %s (%s): transmit timed out\n",
>                                       dev->name, netdev_drivername(dev, drivername, 64));
>                                dev->tx_timeout(dev);
> -                               WARN_ON_ONCE(1);
>                        }
>                        if (!mod_timer(&dev->watchdog_timer,
>                                       round_jiffies(jiffies +
>
>
>
>>I should'nt be posting this, due to
>>having  tainted modules(madwifi), but
>>I figured I might as well:(below)
>>
>>[ 5000.986044] NETDEV WATCHDOG: wifi0: transmit timed out
>>[ 5000.991959] ------------[ cut here ]------------
>>[ 5000.991964] WARNING: at net/sched/sch_generic.c:222 dev_watchdog+0x95/0xe7()
>
> Best Regards
> Komuro
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

Cool thanks for the patch.
I'll plug it in and see if I get this message again.

-- 
Justin P. Mattock

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Arjan van de Ven]

On Tue, 29 Jul 2008 21:53:15 +0900
Komuro <komurojun-mbn@nifty.com> wrote:

> Hi,
> 
> Please try to comment out the "WARN_ON_ONCE"
> at net/sched/sched_generic.c.
> 

"I have a hardware failure so I'll comment out the message that says
so, and then I won't have a failure". Yeah that'll work well.

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[David Miller]

From: Arjan van de Ven <arjan@infradead.org>
Date: Tue, 29 Jul 2008 06:58:59 -0700

> On Tue, 29 Jul 2008 21:53:15 +0900
> Komuro <komurojun-mbn@nifty.com> wrote:
> 
> > Hi,
> > 
> > Please try to comment out the "WARN_ON_ONCE"
> > at net/sched/sched_generic.c.
> > 
> 
> "I have a hardware failure so I'll comment out the message that says
> so, and then I won't have a failure". Yeah that'll work well.

Yes, this is indeed a recidulious suggestion, to hide the problem
instead of trying to figure out what it might actually be.

Komuro has a serious problem accepting the fact that this condition is
in fact an abnormal error should should be logged, and he keeps
pushing to get this WARN_ON() removed over and over again.

I'm just ignoring him at this point.

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Justin Mattock]

On Tue, Jul 29, 2008 at 1:47 PM, David Miller <davem@davemloft.net> wrote:
> From: Arjan van de Ven <arjan@infradead.org>
> Date: Tue, 29 Jul 2008 06:58:59 -0700
>
>> On Tue, 29 Jul 2008 21:53:15 +0900
>> Komuro <komurojun-mbn@nifty.com> wrote:
>>
>> > Hi,
>> >
>> > Please try to comment out the "WARN_ON_ONCE"
>> > at net/sched/sched_generic.c.
>> >
>>
>> "I have a hardware failure so I'll comment out the message that says
>> so, and then I won't have a failure". Yeah that'll work well.
>
> Yes, this is indeed a recidulious suggestion, to hide the problem
> instead of trying to figure out what it might actually be.
>
> Komuro has a serious problem accepting the fact that this condition is
> in fact an abnormal error should should be logged, and he keeps
> pushing to get this WARN_ON() removed over and over again.
>
> I'm just ignoring him at this point.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

I should of mentioned the madwifi module I've been
using is from february, so maybe madwifi took care of this already.
As for experimanting with ath9k I pulled yesterday and compiled,
but from what I saw ath9k seems to not handle wpa keys that well.
so back to madwifi until later on.

-- 
Justin P. Mattock

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Dan Williams]

On Tue, 2008-07-29 at 15:48 -0700, Justin Mattock wrote:
> On Tue, Jul 29, 2008 at 1:47 PM, David Miller <davem@davemloft.net> wrote:
> > From: Arjan van de Ven <arjan@infradead.org>
> > Date: Tue, 29 Jul 2008 06:58:59 -0700
> >
> >> On Tue, 29 Jul 2008 21:53:15 +0900
> >> Komuro <komurojun-mbn@nifty.com> wrote:
> >>
> >> > Hi,
> >> >
> >> > Please try to comment out the "WARN_ON_ONCE"
> >> > at net/sched/sched_generic.c.
> >> >
> >>
> >> "I have a hardware failure so I'll comment out the message that says
> >> so, and then I won't have a failure". Yeah that'll work well.
> >
> > Yes, this is indeed a recidulious suggestion, to hide the problem
> > instead of trying to figure out what it might actually be.
> >
> > Komuro has a serious problem accepting the fact that this condition is
> > in fact an abnormal error should should be logged, and he keeps
> > pushing to get this WARN_ON() removed over and over again.
> >
> > I'm just ignoring him at this point.
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> >
> 
> I should of mentioned the madwifi module I've been
> using is from february, so maybe madwifi took care of this already.
> As for experimanting with ath9k I pulled yesterday and compiled,
> but from what I saw ath9k seems to not handle wpa keys that well.
> so back to madwifi until later on.

ath9k certainly should handle WPA keys; the hardware and driver really
don't care what the key type is because all that is taken care of by
mac80211.  What specific issues were you having with ath9k?  There's
some generic mac80211 breakage going on right now while people figure
out how to fix the skb->cb situation that Dave's multiqueue stuff
exposed, which might be screwing up ath9k for you.

Dan

Re: Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Komuro]

>
>"I have a hardware failure so I'll comment out the message that says
>so, and then I won't have a failure". Yeah that'll work well.


O.K. The madwifi should be fixed.
Thanks,


Best Regards
Komuro

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Arjan van de Ven]

On Wed, 30 Jul 2008 20:48:55 +0900 (JST)
Komuro <komurojun-mbn@nifty.com> wrote:

> 
> >
> >"I have a hardware failure so I'll comment out the message that says
> >so, and then I won't have a failure". Yeah that'll work well.
> 
> 
> O.K. The madwifi should be fixed.
> Thanks,

madwifi has many bigger issues; it's been number 1 in the kerneloops
top 10 for months and months now.

Feel free to fix it.. I can't.. it's a proprietary/binary driver.

-- 
If you want to reach me at my work email, use arjan@linux.intel.com
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org

Re: NETDEV WATCHDOG: wifi0: transmit timed out

[Justin Mattock]

On Wed, Jul 30, 2008 at 6:22 AM, Arjan van de Ven <arjan@infradead.org> wrote:
> On Wed, 30 Jul 2008 20:48:55 +0900 (JST)
> Komuro <komurojun-mbn@nifty.com> wrote:
>
>>
>> >
>> >"I have a hardware failure so I'll comment out the message that says
>> >so, and then I won't have a failure". Yeah that'll work well.
>>
>>
>> O.K. The madwifi should be fixed.
>> Thanks,
>
> madwifi has many bigger issues; it's been number 1 in the kerneloops
> top 10 for months and months now.
>
> Feel free to fix it.. I can't.. it's a proprietary/binary driver.
>
> --
> If you want to reach me at my work email, use arjan@linux.intel.com
> For development, discussion and tips for power savings,
> visit http://www.lesswatts.org
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

I should of written down the error that it was
giving(I can pull again to get the exact error)
from what I remember it was able to do the handshake
but wpa_supplicant would say CTRL-EVENT-CONNECTED then
CTRL-EVENT-DISCONNECTED
if its possible I would like to pull the wireless testing directory
and add that to 2.6.27-rc1. this way I can run ath9k, and the fresh
git "hot out of the oven",
I do have another computer now so reporting problems
makes it a whole lot easier than having one computer.

-- 
Justin P. Mattock