On Tue, Sep 01, 2009 at 06:21:27PM -0700, David Miller wrote:

From: Neil Horman <nhorman@tuxdriver.com>
Date: Mon, 31 Aug 2009 15:58:47 -0400

quoted

It was recently pointed out to me that the last_rx field of the net_device
structure wasn't updated regularly.  In fact only the bonding driver really uses
it currently.  Since the drop_monitor code relies on the last_rx field to detect
drops on recevie in hardware, We need to find a more reliable way to rate limit
our drop checks (so that we don't check for drops on every frame recevied, which
would be inefficient.  This patch makes a last_rx timestamp that is private to
the drop monitor code and is updated for every device that we track.

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>

Neil, this doesn't apply to net-next-2.6:

quoted

diff --git a/net/core/drop_monitor.c b/net/core/drop_monitor.c
index 9d66fa9..34a05ce 100644
--- a/net/core/drop_monitor.c
+++ b/net/core/drop_monitor.c

 ...

quoted

@@ -179,18 +180,21 @@ static void trace_napi_poll_hit(struct napi_struct *napi)
 {
 	struct dm_hw_stat_delta *new_stat;
 
-	/*
-	 * Ratelimit our check time to dm_hw_check_delta jiffies
-	 */
-	if (!time_after(jiffies, napi->dev->last_rx + dm_hw_check_delta))
-		return;
 
 	rcu_read_lock();
 	list_for_each_entry_rcu(new_stat, &hw_stats_list, list) {

In net-next-2.6 this test reads:

	/*
	 * Ratelimit our check time to dm_hw_check_delta jiffies
	 */
	if (!napi->dev ||
	    !time_after(jiffies, napi->dev->last_rx + dm_hw_check_delta))
		return;

and you must retain the napi->dev NULL check there as otherwise
the list traversal tests will blindly dereference that pointer.

Apologies, this raced with Xiao fix to trace_napi_poll hit, which introduced
that null check.  I'll rediff/repost shortly.

Neil