diff mbox

[RFC] csum experts, csum_replace2() is too expensive

Message ID 1395409625.6441.4.camel@edumazet-glaptop2.roam.corp.google.com
State RFC, archived
Delegated to: David Miller
Headers show

Commit Message

Eric Dumazet March 21, 2014, 1:47 p.m. UTC 
On Fri, 2014-03-21 at 06:32 -0700, Eric Dumazet wrote:
> On Fri, 2014-03-21 at 05:50 -0700, Eric Dumazet wrote:
> 
> > Or the fact that we mix 16 bit stores and 32bit loads ?
> > 
> > iph->tot_len = newlen;
> > iph->check = 0;
> > iph->check = ip_fast_csum(iph, 5);
> 
> Yep definitely. Using 16 bit loads in ip_fast_csum() totally removes the
> stall. I no longer see inet_gro_complete() in perf top...
> 
> +       if (__builtin_constant_p(ihl) && ihl == 5) {
> +               asm("  movw (%[iph]), %w[sum]\n"        /* ihl/version/tos */
> +                   "  addw 2(%[iph]), %w[sum]\n"       /* tot_len      */
> +                   "  adcw 8(%[iph]), %w[sum]\n"       /* ttl/protocol */
> +                   "  adcw 10(%[iph]), %w[sum]\n"      /* check        */
> +                   "  adcl 4(%[iph]), %[sum]\n"        /* id/frag_off  */
> +                   "  adcl 12(%[iph]), %[sum]\n"       /* saddr        */
> +                   "  adcl 16(%[iph]), %[sum]\n"       /* daddr        */
> +                   "  adcl $0, %[sum]\n"
> +                   : [sum] "=r" (sum)
> +                   : [iph] "r" (iph)
> +                   );
> +               return csum_fold(sum);
> 

Another idea would be to move the ip_fast_csum() call at the end of
inet_gro_complete()

I'll try this :



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Comments

Eric Dumazet March 21, 2014, 1:56 p.m. UTC | #1 
On Fri, 2014-03-21 at 06:47 -0700, Eric Dumazet wrote:

> Another idea would be to move the ip_fast_csum() call at the end of
> inet_gro_complete()
> 
> I'll try this :
> 
> diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
> index 8c54870db792..0ca8f350a532 100644
> --- a/net/ipv4/af_inet.c
> +++ b/net/ipv4/af_inet.c
> @@ -1434,8 +1434,8 @@ static int inet_gro_complete(struct sk_buff *skb, int nhoff)
>         int proto = iph->protocol;
>         int err = -ENOSYS;
>  
> -       csum_replace2(&iph->check, iph->tot_len, newlen);
>         iph->tot_len = newlen;
> +       iph->check = 0;
>  
>         rcu_read_lock();
>         ops = rcu_dereference(inet_offloads[proto]);
> @@ -1447,6 +1447,7 @@ static int inet_gro_complete(struct sk_buff *skb, int nhoff)
>          * inet_gro_receive().
>          */
>         err = ops->callbacks.gro_complete(skb, nhoff + sizeof(*iph));
> +       iph->check = ip_fast_csum((u8 *)iph, 5);
>  
>  out_unlock:
>         rcu_read_unlock();
> 

This doesn't help... same stall.

Looks like the best way is to use some 16 bit loads in ip_fast_csum()
for the ihl=5 case.



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 8c54870db792..0ca8f350a532 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1434,8 +1434,8 @@  static int inet_gro_complete(struct sk_buff *skb, int nhoff)
        int proto = iph->protocol;
        int err = -ENOSYS;
 
-       csum_replace2(&iph->check, iph->tot_len, newlen);
        iph->tot_len = newlen;
+       iph->check = 0;
 
        rcu_read_lock();
        ops = rcu_dereference(inet_offloads[proto]);
@@ -1447,6 +1447,7 @@  static int inet_gro_complete(struct sk_buff *skb, int nhoff)
         * inet_gro_receive().
         */
        err = ops->callbacks.gro_complete(skb, nhoff + sizeof(*iph));
+       iph->check = ip_fast_csum((u8 *)iph, 5);
 
 out_unlock:
        rcu_read_unlock();