Re: [PATCH 2.6.9-rc3] dev_close: use of netif_poll

netdev

Re: [PATCH 2.6.9-rc3] dev_close: use of netif_poll_disable

To:	Francois Romieu <romieu@xxxxxxxxxxxxx>
Subject:	Re: [PATCH 2.6.9-rc3] dev_close: use of netif_poll_disable
From:	"David S. Miller" <davem@xxxxxxxxxx>
Date:	Sun, 3 Oct 2004 15:02:29 -0700
Cc:	netdev@xxxxxxxxxxx
In-reply-to:	<20041002202805.GA16870@xxxxxxxxxxxxxxxxxxxxxxxxxx>
References:	<20041002202805.GA16870@xxxxxxxxxxxxxxxxxxxxxxxxxx>
Sender:	netdev-bounce@xxxxxxxxxxx

On Sat, 2 Oct 2004 22:28:05 +0200
Francois Romieu <romieu@xxxxxxxxxxxxx> wrote:

> Use of identical (comment included) netif_poll_disable().
> 
> Signed-off-by: Francois Romieu <romieu@xxxxxxxxxxxxx>

It's not identical.  netif_poll_disable() actually sets the
__LINK_STATE_RX_SCHED bit when it finally sees it clear.

Whereas this code in net/core/dev.c is just waiting for it
to clear.

[More with this subject...]

<Prev in Thread]	Current Thread	[Next in Thread>
[PATCH 2.6.9-rc3] dev_close: use of netif_poll_disable, Francois Romieu Re: [PATCH 2.6.9-rc3] dev_close: use of netif_poll_disable, David S. Miller <=

Previous by Date:	Re: [PATCH] fix secure tcp sequence number generation, David S. Miller
Next by Date:	Re: Resend WAS(Re: cls patchlet, David S. Miller
Previous by Thread:	[PATCH 2.6.9-rc3] dev_close: use of netif_poll_disable, Francois Romieu
Next by Thread:	i82596 ethernet controller, Arthur Rinkel
Indexes:	[Date] [Thread] [Top] [All Lists]