Re: dev->xmit_lock

netdev

Re: dev->xmit_lock_owner?

To:	jgarzik@xxxxxxxxx
Subject:	Re: dev->xmit_lock_owner?
From:	"David S. Miller" <davem@xxxxxxxxxx>
Date:	Tue, 18 Feb 2003 16:02:41 -0800 (PST)
Cc:	linux-net@xxxxxxxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to:	<20030218183306.GA31478@gtf.org>
References:	<20030218183306.GA31478@gtf.org>
Sender:	netdev-bounce@xxxxxxxxxxx

   From: Jeff Garzik <jgarzik@xxxxxxxxx>
   Date: Tue, 18 Feb 2003 13:33:06 -0500

   It seems that dev_watchdog, among other places, does not assign a value
   to dev->xmit_lock_owner, when it takes the lock.

   I think this is a bug, but could be wrong ;-)

Not a bug, only pieces of the transmit path need to follow
this rule.  It's only meant to detect devices which have
been chained in a way which forms a loop during transmit.

[More with this subject...]

<Prev in Thread]	Current Thread	[Next in Thread>
dev->xmit_lock_owner?, Jeff Garzik Re: dev->xmit_lock_owner?, David S. Miller <=

Previous by Date:	Re: [PATCH resend] zero rt_cache_stat statistics at init, David S. Miller
Next by Date:	Re: [PATCH] subset of RFC2553, Bruce Allan
Previous by Thread:	dev->xmit_lock_owner?, Jeff Garzik
Next by Thread:	netdevices.txt update, Jeff Garzik
Indexes:	[Date] [Thread] [Top] [All Lists]