netdev - Re: Softirqs without captured packets

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Sun, 10 Feb 2013 00:16:28 -0800
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Javier Domingo <javierdo1@...il.com>
Cc:	David Miller <davem@...emloft.net>, netdev@...r.kernel.org
Subject: Re: Softirqs without captured packets

Please do not top post on netdev

On Sun, 2013-02-10 at 04:02 +0100, Javier Domingo wrote:
> Well, that explains why the 0 :D. I was getting nut. Any way, does
> that happen in e1000e or tg3 drivers? If not, then there is still
> without explanation.
> 


The answer is yes.

> Respecting the low numbers, I will do some futher stress tests, but
> this measures where done pinging the computer at 0.001 interval:
> 
> ping -i 0.0001 <computer ip>
> 
> And received packets where at much, 3 per softirq. Also, I tried
> opening an ftp download session but didn't get much high numbers
> (1Gbps link).
> 

Machine is fast enough, thats all.

> When I tried this with the injection done by a dag (hardware injector)
> there were strange results because thought the 0s where explained, the
> were appearing thought the rest of softirqs where about 280+-50.

Have you read this part of net_rx_action() ?

unsigned long time_limit = jiffies + 2;

                /* If softirq window is exhuasted then punt.
                 * Allow this to run for 2 jiffies since which will allow
                 * an average latency of 1.5/HZ.
                 */
                if (unlikely(budget <= 0 || time_after(jiffies, time_limit)))
                        goto softnet_break;

This means your machine handles 280+50 packets per 1.5/HZ

(Plus the hidden TX completions...)



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