Re: [RFC v2] mac80211: assign needed_headroom/tailroom for netdevs

[Johannes Berg <johannes@sipsolutions.net>]

[-- Attachment #1: Type: text/plain, Size: 1552 bytes --]

On Mon, 2008-05-05 at 16:24 -0700, David Miller wrote:

> > Also, should there be some sort of timer that resets the rx_alloc_extra
> > again so that when you bridge it once with p54 (needs heaps of headroom)
> > you don't suffer forever?
> 
> We're talking about, what, up to 128 bytes or something like that?
> If you're bridging over your wireless device you've already invested
> in assuming those kinds of costs.

About that much at most, yeah. But if you then decide that you don't
want to bridge wireless after all... anyway doesn't matter, was just a
thought.

> I don't think this aspect is really worth worrying about.  The current
> behavior is so much incredibly worse. :-)

That's true.

> > > @@ -255,11 +255,12 @@ struct sk_buff *__netdev_alloc_skb(struct net_device *dev,
> > >  		unsigned int length, gfp_t gfp_mask)
> > >  {
> > >  	int node = dev->dev.parent ? dev_to_node(dev->dev.parent) : -1;
> > > +	unsigned int extra = dev->rx_alloc_extra + NET_SKB_PAD;
> > >  	struct sk_buff *skb;
> > >  
> > > -	skb = __alloc_skb(length + NET_SKB_PAD, gfp_mask, 0, node);
> > > +	skb = __alloc_skb(length + extra, gfp_mask, 0, node);
> > >  	if (likely(skb)) {
> > > -		skb_reserve(skb, NET_SKB_PAD);
> > > +		skb_reserve(skb, extra);
> > 
> > Doesn't that break alignment though?
> 
> Good catch, we'll have to align the rx_alloc_extra to some modulus or
> similar.  Ideally, at the spot where rx_alloc_extra is set instead
> of here.

Yeah, it just has to be a multiple of four I think.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]