2024-02-18 18:36:09

by Oliver Crumrine

Subject: [PATCH net-next] net: copy routing table more efficiently in rt_dst_clone

Instead of copying field-by-field, copy the entire struct at once. This
should lead to a performance improvment in rt_dst_clone.

Signed-off-by: Oliver Crumrine <[email protected]>
---
net/ipv4/route.c | 16 +---------------
1 file changed, 1 insertion(+), 15 deletions(-)

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 16615d107cf0..ebb17c3a0dec 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -1664,22 +1664,8 @@ struct rtable *rt_dst_clone(struct net_device *dev, struct rtable *rt)
rt->dst.flags);

if (new_rt) {
+ *new_rt = *rt;
new_rt->rt_genid = rt_genid_ipv4(dev_net(dev));
- new_rt->rt_flags = rt->rt_flags;
- new_rt->rt_type = rt->rt_type;
- new_rt->rt_is_input = rt->rt_is_input;
- new_rt->rt_iif = rt->rt_iif;
- new_rt->rt_pmtu = rt->rt_pmtu;
- new_rt->rt_mtu_locked = rt->rt_mtu_locked;
- new_rt->rt_gw_family = rt->rt_gw_family;
- if (rt->rt_gw_family == AF_INET)
- new_rt->rt_gw4 = rt->rt_gw4;
- else if (rt->rt_gw_family == AF_INET6)
- new_rt->rt_gw6 = rt->rt_gw6;
-
- new_rt->dst.input = rt->dst.input;
- new_rt->dst.output = rt->dst.output;
- new_rt->dst.error = rt->dst.error;
new_rt->dst.lastuse = jiffies;
new_rt->dst.lwtstate = lwtstate_get(rt->dst.lwtstate);
}
--
2.43.0

2024-02-21 02:44:36

by David Ahern

Subject: Re: [PATCH net-next] net: copy routing table more efficiently in rt_dst_clone

On 2/18/24 6:35 AM, Oliver Crumrine wrote:
> diff --git a/net/ipv4/route.c b/net/ipv4/route.c
> index 16615d107cf0..ebb17c3a0dec 100644
> --- a/net/ipv4/route.c
> +++ b/net/ipv4/route.c
> @@ -1664,22 +1664,8 @@ struct rtable *rt_dst_clone(struct net_device *dev, struct rtable *rt)
> rt->dst.flags);
>
> if (new_rt) {
> + *new_rt = *rt;

rtable is a container of dst_entry, so this is copying those fields as well.

pw-bot: reject

2024-02-21 23:52:42

by Oliver Crumrine

Subject: Re: [PATCH net-next] net: copy routing table more efficiently in rt_dst_clone

On Tue, Feb 20, 2024 at 07:44:24PM -0700, David Ahern wrote:
Hi David,

You are correct that rtable is a container of dst_entry, and the
previous code copied not only the fields of rtable but also the fields
of dst_entry. However, the *new_rt = *rt line not only copies the fields
of rtable, but also the fields of dst_entry.

I have demonstrated this by generating a random number and putting it in
the "expired" field of the dst_entry that is part of rt, and the printing
the random number that was generated. After the copy, I printed the
expired field of the dst_entry that is part of new_rt and the numbers
matched. This proves that not only the fields that were part of rtable
were copied but also the fields of dst_entry.

Thanks,
Oliver
> On 2/18/24 6:35 AM, Oliver Crumrine wrote:
> > diff --git a/net/ipv4/route.c b/net/ipv4/route.c
> > index 16615d107cf0..ebb17c3a0dec 100644
> > --- a/net/ipv4/route.c
> > +++ b/net/ipv4/route.c
> > @@ -1664,22 +1664,8 @@ struct rtable *rt_dst_clone(struct net_device *dev, struct rtable *rt)
> > rt->dst.flags);
> >
> > if (new_rt) {
> > + *new_rt = *rt;
>
> rtable is a container of dst_entry, so this is copying those fields as well.
>
> pw-bot: reject
>