Re: [PATCH net-next 08/17] netdevsim: Adjust accounting for IPv6 multipath notifications

[Ido Schimmel <idosch@idosch.org>]

On Sun, Jun 16, 2019 at 07:27:13PM -0600, David Ahern wrote:
> On 6/15/19 8:07 AM, Ido Schimmel wrote:
> > diff --git a/drivers/net/netdevsim/fib.c b/drivers/net/netdevsim/fib.c
> > index 83ba5113210d..6e5498ef3855 100644
> > --- a/drivers/net/netdevsim/fib.c
> > +++ b/drivers/net/netdevsim/fib.c
> > @@ -137,19 +137,20 @@ static int nsim_fib_rule_event(struct nsim_fib_data *data,
> >  }
> >  
> >  static int nsim_fib_account(struct nsim_fib_entry *entry, bool add,
> > +			    unsigned int num_rt,
> >  			    struct netlink_ext_ack *extack)
> >  {
> >  	int err = 0;
> >  
> >  	if (add) {
> > -		if (entry->num < entry->max) {
> > -			entry->num++;
> > +		if (entry->num + num_rt < entry->max) {
> > +			entry->num += num_rt;
> >  		} else {
> >  			err = -ENOSPC;
> >  			NL_SET_ERR_MSG_MOD(extack, "Exceeded number of supported fib entries");
> >  		}
> >  	} else {
> > -		entry->num--;
> > +		entry->num -= num_rt;
> >  	}
> >  
> >  	return err;
> > @@ -159,14 +160,20 @@ static int nsim_fib_event(struct nsim_fib_data *data,
> >  			  struct fib_notifier_info *info, bool add)
> >  {
> >  	struct netlink_ext_ack *extack = info->extack;
> > +	struct fib6_entry_notifier_info *fen6_info;
> > +	unsigned int num_rt = 1;
> >  	int err = 0;
> >  
> >  	switch (info->family) {
> >  	case AF_INET:
> > -		err = nsim_fib_account(&data->ipv4.fib, add, extack);
> > +		err = nsim_fib_account(&data->ipv4.fib, add, num_rt, extack);
> >  		break;
> >  	case AF_INET6:
> > -		err = nsim_fib_account(&data->ipv6.fib, add, extack);
> > +		fen6_info = container_of(info, struct fib6_entry_notifier_info,
> > +					 info);
> > +		if (fen6_info->multipath_rt)
> > +			num_rt = fen6_info->nsiblings + 1;
> > +		err = nsim_fib_account(&data->ipv6.fib, add, num_rt, extack);
> 
> The intention of the original patch was to account for a multipath route
> as 1 entry, not N where N is the number of paths.

:)

Didn't want to change existing behavior in case you have tests that
would fail because of that. I guess that's not the case. Will treat it
as one entry.