From mboxrd@z Thu Jan  1 00:00:00 1970
From: Eric Dumazet <dada1@cosmosbay.com>
Subject: Re: [PATCH 2/2] udp: RCU handling for Unicast packets.
Date: Thu, 30 Oct 2008 08:04:42 +0100
Message-ID: <49095C8A.403@cosmosbay.com>
References: <4908C0CD.5050406@cosmosbay.com>	<20081029201759.GF6732@linux.vnet.ibm.com>	<4908DEDE.5030706@cosmosbay.com> <20081029.224027.163185414.davem@davemloft.net> <49094B79.4010805@cosmosbay.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1;
	format=flowed
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: paulmck@linux.vnet.ibm.com, minyard@acm.org, shemminger@vyatta.com,
	benny+usenet@amorsen.dk, netdev@vger.kernel.org,
	cl@linux-foundation.org, a.p.zijlstra@chello.nl,
	johnpol@2ka.mipt.ru, christian@myri.com
To: David Miller <davem@davemloft.net>
Return-path: <netdev-owner@vger.kernel.org>
Received: from gw1.cosmosbay.com ([86.65.150.130]:41760 "EHLO
	gw1.cosmosbay.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752891AbYJ3HFR convert rfc822-to-8bit (ORCPT
	<rfc822;netdev@vger.kernel.org>); Thu, 30 Oct 2008 03:05:17 -0400
In-Reply-To: <49094B79.4010805@cosmosbay.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

Eric Dumazet a =E9crit :
> David Miller a =E9crit :
>> From: Eric Dumazet <dada1@cosmosbay.com>
>> Date: Wed, 29 Oct 2008 23:08:30 +0100
>>
>>> @@ -1746,7 +1753,7 @@ void __init udp_table_init(struct udp_table=20
>>> *table)
>>>      int i;
>>> =20
>>>      for (i =3D 0; i < UDP_HTABLE_SIZE; i++) {
>>> -        INIT_HLIST_HEAD(&table->hash[i].head);
>>> +        table->hash[i].head.first =3D (struct hlist_node *)i;
>>
>> Please hide this behind some list.h interface macro, even something
>> as simple as INIT_HLIST_HEAD_NULLS(X, index) would suffice.
>>
>> And as Corey said, the code needs more comments for something as
>> clever as this! :-)
>>
>=20
> Yes I agree 100%, please give me one day to prepare a real patch,
> or else akpm will kill us :)
>=20

If we design something that could be reused, say for TCP sockets, we ne=
ed
to be able to handle very large number of 'NULL' pointers, say, up to 6=
4*1024*1024

So lets use the low order bit, set to one for "NULL" pointers, and 0 fo=
r regular pointers.

This gives us 31 bits (or 63 bits) to store any valuable info :)

and all ...._nulls() macros would not need to know the max value (UDP_H=
TABLE_SIZE in UDP case),
since all they have to do is a test ((unsigned long)pos & 1)

At iterator exit, pos would contain the 'index' value, (pos >> 1), to h=
ide this
implementation detail.