netdev - Re: [PATCH net-next 2/3] inet: get rid of central tcp/dccp listener timer

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:	Thu, 19 Mar 2015 19:41:59 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Alexei Starovoitov <alexei.starovoitov@...il.com>
Cc:	Eric Dumazet <edumazet@...gle.com>,
	"David S. Miller" <davem@...emloft.net>,
	netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next 2/3] inet: get rid of central tcp/dccp listener
 timer

On Thu, 2015-03-19 at 19:26 -0700, Alexei Starovoitov wrote:

> > This is ~100 times more what we could achieve before this patch.
> 
> wow. That is impressive!
> I guess timers do scale to millions.
> Outstanding.

Note that my goal is go get 1000 times, when we no longer take listener
lock ;)

When req sock is hashed into ehash, this means the ACK packet coming
from the client will find it while doing the normal RCU lookup.

ehash is protected by an array of spinlocks, so it scales.

Ideally, we should allow say 10 million request socks, so that SYNCOOKIE
mode is only a real emergency thing.

Thanks.


--
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