Re: [PATCH] decouple llc/bridge

[Stephen Hemminger <stephen@networkplumber.org>]

On Thu, 7 Apr 2022 19:48:59 -0700
Jakub Kicinski <kuba@kernel.org> wrote:

> On Thu, 7 Apr 2022 09:16:40 -0700 Stephen Hemminger wrote:
> > > I was wondering why the llc code was getting compiled and it turned out
> > > to be because I had bridging enabled. It turns out to only needs it for
> > > a single function (llc_mac_hdr_init).  
> 
> > > +static inline int llc_mac_hdr_init(struct sk_buff *skb,
> > > +				   const unsigned char *sa, const unsigned char *da)
> > > +{
> > > +	int rc = -EINVAL;
> > > +
> > > +	switch (skb->dev->type) {
> > > +	case ARPHRD_ETHER:
> > > +	case ARPHRD_LOOPBACK:
> > > +		rc = dev_hard_header(skb, skb->dev, ETH_P_802_2, da, sa,
> > > +				     skb->len);
> > > +		if (rc > 0)
> > > +			rc = 0;
> > > +		break;
> > > +	default:
> > > +		break;
> > > +	}
> > > +	return rc;
> > > +}
> > > +
> > >    
> 
> nit: extra new line
> 
> > > -int llc_mac_hdr_init(struct sk_buff *skb,
> > > -		     const unsigned char *sa, const unsigned char *da)
> > > -{
> > > -	int rc = -EINVAL;
> > > -
> > > -	switch (skb->dev->type) {
> > > -	case ARPHRD_ETHER:
> > > -	case ARPHRD_LOOPBACK:
> > > -		rc = dev_hard_header(skb, skb->dev, ETH_P_802_2, da, sa,
> > > -				     skb->len);
> > > -		if (rc > 0)
> > > -			rc = 0;
> > > -		break;
> > > -	default:
> > > -		break;
> > > -	}
> > > -	return rc;
> > > -}  
> 
> There's also an EXPORT somewhere in this file that has to go.
> 
> > >  /**
> > >   *	llc_build_and_send_ui_pkt - unitdata request interface for upper layers
> > >   *	@sap: sap to use    
> > 
> > You may break other uses of LLC.
> > 
> > Why not open code as different function.  I used the llc stuff because there
> > were multiple copies of same code (DRY).  
> 
> I didn't quite get what you mean, Stephen, would you mind restating?


Short answer: it was idea that didn't pan out.

Suggestion: get rid of using LLC  in bridge and just rewrite that one place.