[PATCH 1/2] netprio_cgroup: Remove update_netdev_tables() since it is unnecessary

[PATCH 1/2] netprio_cgroup: Remove update_netdev_tables() since it is unnecessary

[Srivatsa S. Bhat]

The update_netdev_tables() function appears to be unnecessary, since the
write_update_netdev_table() function will adjust the priomaps as and when
required anyway. So drop the usage of update_netdev_tables() entirely.

Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---

 net/core/netprio_cgroup.c |   32 --------------------------------
 1 files changed, 0 insertions(+), 32 deletions(-)

diff --git a/net/core/netprio_cgroup.c b/net/core/netprio_cgroup.c
index c75e3f9..fd339bb0 100644
--- a/net/core/netprio_cgroup.c
+++ b/net/core/netprio_cgroup.c
@@ -109,32 +109,6 @@ static int write_update_netdev_table(struct net_device *dev)
 	return ret;
 }
 
-static int update_netdev_tables(void)
-{
-	int ret = 0;
-	struct net_device *dev;
-	u32 max_len;
-	struct netprio_map *map;
-
-	rtnl_lock();
-	max_len = atomic_read(&max_prioidx) + 1;
-	for_each_netdev(&init_net, dev) {
-		map = rtnl_dereference(dev->priomap);
-		/*
-		 * don't allocate priomap if we didn't
-		 * change net_prio.ifpriomap (map == NULL),
-		 * this will speed up skb_update_prio.
-		 */
-		if (map && map->priomap_len < max_len) {
-			ret = extend_netdev_table(dev, max_len);
-			if (ret < 0)
-				break;
-		}
-	}
-	rtnl_unlock();
-	return ret;
-}
-
 static struct cgroup_subsys_state *cgrp_create(struct cgroup *cgrp)
 {
 	struct cgroup_netprio_state *cs;
@@ -153,12 +127,6 @@ static struct cgroup_subsys_state *cgrp_create(struct cgroup *cgrp)
 		goto out;
 	}
 
-	ret = update_netdev_tables();
-	if (ret < 0) {
-		put_prioidx(cs->prioidx);
-		goto out;
-	}
-
 	return &cs->css;
 out:
 	kfree(cs);

[PATCH 2/2] netprio_cgroup: Optimize the priomap copy loop slightly

[Srivatsa S. Bhat]

* Check for non-NULL old_priomap outside the loop, since its
  not going to change.
* Copy the old_priomap's length to a local variable and use
  that for loop control, instead of costly pointer-dereferences.

Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---

 net/core/netprio_cgroup.c |   12 +++++++-----
 1 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/net/core/netprio_cgroup.c b/net/core/netprio_cgroup.c
index fd339bb0..0775bc9 100644
--- a/net/core/netprio_cgroup.c
+++ b/net/core/netprio_cgroup.c
@@ -73,7 +73,7 @@ static int extend_netdev_table(struct net_device *dev, u32 new_len)
 			   ((sizeof(u32) * new_len));
 	struct netprio_map *new_priomap = kzalloc(new_size, GFP_KERNEL);
 	struct netprio_map *old_priomap;
-	int i;
+	int i, old_len;
 
 	old_priomap  = rtnl_dereference(dev->priomap);
 
@@ -82,10 +82,12 @@ static int extend_netdev_table(struct net_device *dev, u32 new_len)
 		return -ENOMEM;
 	}
 
-	for (i = 0;
-	     old_priomap && (i < old_priomap->priomap_len);
-	     i++)
-		new_priomap->priomap[i] = old_priomap->priomap[i];
+	if (old_priomap) {
+		old_len = old_priomap->priomap_len;
+
+		for (i = 0; i < old_len; i++)
+			new_priomap->priomap[i] = old_priomap->priomap[i];
+	}
 
 	new_priomap->priomap_len = new_len;
 

RE: [PATCH 2/2] netprio_cgroup: Optimize the priomap copy loop slightly

[David Laight]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="utf-8", Size: 596 bytes --]

> -	for (i = 0;
> -	     old_priomap && (i < old_priomap->priomap_len);
> -	     i++)
> -		new_priomap->priomap[i] = old_priomap->priomap[i];
> +	if (old_priomap) {
> +		old_len = old_priomap->priomap_len;
> +
> +		for (i = 0; i < old_len; i++)
> +			new_priomap->priomap[i] = old_priomap->priomap[i];
> +	}

Or:
	memcpy(new_priomap->priomap, old_priomap->priomap,
		old_priomap->priomap_len * sizeof old_priomap->priomap[0]);

	David

ÿôèº{.nÇ+‰·Ÿ®‰­†+%ŠËÿ±éݶ\x17¥Šwÿº{.nÇ+‰·¥Š{±þG«éÿŠ{ayº\x1dʇڙë,j\a­¢f£¢·hšïêÿ‘êçz_è®\x03(­éšŽŠÝ¢j"ú\x1a¶^[m§ÿÿ¾\a«þG«éÿ¢¸?™¨è­Ú&£ø§~á¶iO•æ¬z·švØ^\x14\x04\x1a¶^[m§ÿÿÃ\fÿ¶ìÿ¢¸?–I¥

Re: [PATCH 2/2] netprio_cgroup: Optimize the priomap copy loop slightly

[Neil Horman]

On Tue, Sep 11, 2012 at 12:42:23PM +0100, David Laight wrote:
> > -	for (i = 0;
> > -	     old_priomap && (i < old_priomap->priomap_len);
> > -	     i++)
> > -		new_priomap->priomap[i] = old_priomap->priomap[i];
> > +	if (old_priomap) {
> > +		old_len = old_priomap->priomap_len;
> > +
> > +		for (i = 0; i < old_len; i++)
> > +			new_priomap->priomap[i] = old_priomap->priomap[i];
> > +	}
> 
> Or:
> 	memcpy(new_priomap->priomap, old_priomap->priomap,
> 		old_priomap->priomap_len * sizeof old_priomap->priomap[0]);
> 
> 	David
> 
Yes, the memcpy would be better here.
Neil

Re: [PATCH 2/2] netprio_cgroup: Optimize the priomap copy loop slightly

[Srivatsa S. Bhat]

On 09/11/2012 05:12 PM, David Laight wrote:
>> -	for (i = 0;
>> -	     old_priomap && (i < old_priomap->priomap_len);
>> -	     i++)
>> -		new_priomap->priomap[i] = old_priomap->priomap[i];
>> +	if (old_priomap) {
>> +		old_len = old_priomap->priomap_len;
>> +
>> +		for (i = 0; i < old_len; i++)
>> +			new_priomap->priomap[i] = old_priomap->priomap[i];
>> +	}
> 
> Or:
> 	memcpy(new_priomap->priomap, old_priomap->priomap,
> 		old_priomap->priomap_len * sizeof old_priomap->priomap[0]);
> 

Ah, indeed that would be better. I'll send out an updated version of the
patches. Thanks!

Regards,
Srivatsa S. Bhat