[PATCH] netpoll: fix TX queue overflow in trapped mode

[PATCH] netpoll: fix TX queue overflow in trapped mode

[Sergei Shtylyov]

CONFIG_NETPOLL_TRAP causes the TX queue controls to be completely bypassed in
the netpoll's "trapped" mode which easily causes overflows in the drivers with
short TX queues (most notably, in 8139too with its 4-deep queue).  So, make
this option more sensible by making it only bypass the TX softirq wakeup.

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>

---
Split this part form the initial patch as request by Matt Mackall.

 include/linux/netdevice.h |    8 +++-----
 2 files changed, 3 insertions(+), 10 deletions(-)

Index: linux-2.6/include/linux/netdevice.h
===================================================================
--- linux-2.6.orig/include/linux/netdevice.h
+++ linux-2.6/include/linux/netdevice.h
@@ -647,8 +647,10 @@ static inline void netif_start_queue(str
 static inline void netif_wake_queue(struct net_device *dev)
 {
 #ifdef CONFIG_NETPOLL_TRAP
-	if (netpoll_trap())
+	if (netpoll_trap()) {
+		clear_bit(__LINK_STATE_XOFF, &dev->state);
 		return;
+	}
 #endif
 	if (test_and_clear_bit(__LINK_STATE_XOFF, &dev->state))
 		__netif_schedule(dev);
@@ -656,10 +658,6 @@ static inline void netif_wake_queue(stru
 
 static inline void netif_stop_queue(struct net_device *dev)
 {
-#ifdef CONFIG_NETPOLL_TRAP
-	if (netpoll_trap())
-		return;
-#endif
 	set_bit(__LINK_STATE_XOFF, &dev->state);
 }
 

Re: [PATCH] netpoll: fix TX queue overflow in trapped mode

[Sergei Shtylyov]

On Saturday 28 April 2007 22:12, you wrote:
> CONFIG_NETPOLL_TRAP causes the TX queue controls to be completely bypassed
> in the netpoll's "trapped" mode which easily causes overflows in the
> drivers with short TX queues (most notably, in 8139too with its 4-deep
> queue).  So, make this option more sensible by making it only bypass the TX
> softirq wakeup.

   Dammit!  Finally forgot to add acks earned so far:

> Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Acked-by: Jeff Garzik <jgarzik@pobox.com>
Acked-by: Matt Mackall <mpm@selenic.com>

> ---
> Split this part form the initial patch as request by Matt Mackall.
>
>  include/linux/netdevice.h |    8 +++-----
>  2 files changed, 3 insertions(+), 10 deletions(-)
>
> Index: linux-2.6/include/linux/netdevice.h
> ===================================================================
> --- linux-2.6.orig/include/linux/netdevice.h
> +++ linux-2.6/include/linux/netdevice.h
> @@ -647,8 +647,10 @@ static inline void netif_start_queue(str
>  static inline void netif_wake_queue(struct net_device *dev)
>  {
>  #ifdef CONFIG_NETPOLL_TRAP
> -	if (netpoll_trap())
> +	if (netpoll_trap()) {
> +		clear_bit(__LINK_STATE_XOFF, &dev->state);
>  		return;
> +	}
>  #endif
>  	if (test_and_clear_bit(__LINK_STATE_XOFF, &dev->state))
>  		__netif_schedule(dev);
> @@ -656,10 +658,6 @@ static inline void netif_wake_queue(stru
>
>  static inline void netif_stop_queue(struct net_device *dev)
>  {
> -#ifdef CONFIG_NETPOLL_TRAP
> -	if (netpoll_trap())
> -		return;
> -#endif
>  	set_bit(__LINK_STATE_XOFF, &dev->state);
>  }

Re: [PATCH] netpoll: fix TX queue overflow in trapped mode

[Sergei Shtylyov]

On Saturday 28 April 2007 22:52, I wrote:

> > CONFIG_NETPOLL_TRAP causes the TX queue controls to be completely
> > bypassed in the netpoll's "trapped" mode which easily causes overflows in
> > the drivers with short TX queues (most notably, in 8139too with its
> > 4-deep queue).  So, make this option more sensible by making it only
> > bypass the TX softirq wakeup.
>    Dammit!  Finally forgot to add acks earned so far:

   Oh, and there was even another one:

> > Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>

> Acked-by: Jeff Garzik <jgarzik@pobox.com>
Acked-by: Tom Rini <trini@kernel.crashing.org>
> Acked-by: Matt Mackall <mpm@selenic.com>

> > ---
> > Split this part form the initial patch as request by Matt Mackall.
> >
> >  include/linux/netdevice.h |    8 +++-----
> >  2 files changed, 3 insertions(+), 10 deletions(-)
> >
> > Index: linux-2.6/include/linux/netdevice.h
> > ===================================================================
> > --- linux-2.6.orig/include/linux/netdevice.h
> > +++ linux-2.6/include/linux/netdevice.h
> > @@ -647,8 +647,10 @@ static inline void netif_start_queue(str
> >  static inline void netif_wake_queue(struct net_device *dev)
> >  {
> >  #ifdef CONFIG_NETPOLL_TRAP
> > -	if (netpoll_trap())
> > +	if (netpoll_trap()) {
> > +		clear_bit(__LINK_STATE_XOFF, &dev->state);
> >  		return;
> > +	}
> >  #endif
> >  	if (test_and_clear_bit(__LINK_STATE_XOFF, &dev->state))
> >  		__netif_schedule(dev);
> > @@ -656,10 +658,6 @@ static inline void netif_wake_queue(stru
> >
> >  static inline void netif_stop_queue(struct net_device *dev)
> >  {
> > -#ifdef CONFIG_NETPOLL_TRAP
> > -	if (netpoll_trap())
> > -		return;
> > -#endif
> >  	set_bit(__LINK_STATE_XOFF, &dev->state);
> >  }

Re: [PATCH] netpoll: fix TX queue overflow in trapped mode

[David Miller]

From: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Date: Sun, 29 Apr 2007 00:25:41 +0400

> On Saturday 28 April 2007 22:52, I wrote:
> 
> > > CONFIG_NETPOLL_TRAP causes the TX queue controls to be completely
> > > bypassed in the netpoll's "trapped" mode which easily causes overflows in
> > > the drivers with short TX queues (most notably, in 8139too with its
> > > 4-deep queue).  So, make this option more sensible by making it only
> > > bypass the TX softirq wakeup.
> >    Dammit!  Finally forgot to add acks earned so far:
> 
>    Oh, and there was even another one:
> 
> > > Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
> 
> > Acked-by: Jeff Garzik <jgarzik@pobox.com>
> Acked-by: Tom Rini <trini@kernel.crashing.org>
> > Acked-by: Matt Mackall <mpm@selenic.com>

Applied, I'll push this to -stable too.