netdev - Re: many sockets, slow sendto

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

Message-Id: <20070430144715.b0c03c83.dada1@cosmosbay.com>
Date:	Mon, 30 Apr 2007 14:47:15 +0200
From:	Eric Dumazet <dada1@...mosbay.com>
To:	YOSHIFUJI Hideaki / 吉藤英明 
	<yoshfuji@...ux-ipv6.org>
Cc:	davem@...emloft.net, zacco@...hu, baruch@...en.org,
	netdev@...r.kernel.org
Subject: Re: many sockets, slow sendto

On Mon, 30 Apr 2007 19:56:04 +0900 (JST)
YOSHIFUJI Hideaki / 吉藤英明 <yoshfuji@...ux-ipv6.org> wrote:

> In article <20070430.002643.68156452.davem@...emloft.net> (at Mon, 30 Apr 2007 00:26:43 -0700 (PDT)), David Miller <davem@...emloft.net> says:
> 
> > > Signed-off-by: Eric Dumazet <dada1@...mosbay.com>
> > 
> > Eric, I've applied this, thanks again.
> > 
> > Could I trouble you to cook up an ipv6 version of this patch?
> 
> Here's my tentative version.  Not tested.
> Dave, Eric, could you double-check this, please?

Hi Yoshifuji

Your patch seems good, thank you.

I have two minor comments about udp6_hash_port_and_addr()

Compiler knows that const 'struct in6_addr *addr' is NULL when we call udp6_hash_port_and_addr(..., NULL)

But not when we pass saddr :  udp6_hash_port_and_addr(port, saddr)

So to avoid one conditional branch, you could use udp6_hash_port_and_addr() only once, with addr not NULL :

Also, I am not sure we need to use all 128 bits of IPV6 address, maybe the 64 low order bits are enough ?

static inline unsigned int udp6_hash_port_and_addr(__u16 port,
						  const struct in6_addr *addr)
{
	u32 hash = (__force u32) addr->s6_addr32[2] ^
		   (__force u32) addr->s6_addr32[3];
	hash ^= hash >> 16;
	hash ^= hash >> 8;

	return port ^ hash;
}

static inline unsigned int udp6_hash_wildcard(__u16 port,
					      const struct in6_addr *addr)
{
	return port;
}

...


	hash = udp6_hash_port_and_addr(ntohs(dport), saddr);
	hashwild = udp6_hash_wildcard(ntohs(dport), saddr);

Eric
-
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