From: Patrick McHardy <kaber@trash.net>
To: Jan Engelhardt <jengelh@computergmbh.de>
Cc: Netfilter Developer Mailing List <netfilter-devel@lists.netfilter.org>
Subject: Re: xt_connlimit 20070625
Date: Mon, 25 Jun 2007 16:46:06 +0200 [thread overview]
Message-ID: <467FD52E.4050705@trash.net> (raw)
In-Reply-To: <Pine.LNX.4.64.0706251451000.3137@fbirervta.pbzchgretzou.qr>
Jan Engelhardt wrote:
> On Jun 25 2007 14:43, Patrick McHardy wrote:
>
>>>Indeed - leftovers from its POM ascendant. But do I have to add ipv6 now?
>>>
>>>(It would also be quite a duplicate of count_them() and connlimit_match() -
>>>is there a better way without too many if()s?)
>>>
>>
>>I'd use a conditional in count_them and add an connlimit_match6 function.
>
>
> Thing is..
> static int count_them(struct xt_connlimit_data *data, u_int32_t addr,
> u_int32_t mask, struct nf_conn *ct)
>
> I'd need a u_int128_t for IPv6. (Or a clever idea.)
> Suggestions for elegant solution?
struct xt_conntrack_address
>>>>Incorrect comment, it does use &.
>>>
>>>I reword it.
>
>
> /*
> * The hash needs to be folded down (using the % modulo operator) to
> * actually yield a value that is greater or equal than 0 and smaller
> * than ARRAY_SIZE(iphash). If ARRAY_SIZE(iphash) is exactly a power of
> * two, the (most likely faster) "& mask" operation can also be used,
> * where mask is the size minus one.
> */
>
> Does that read better?
That falls into the "i++ /* increment i */" category IMO.
If someone doesn't know how basic operators work he won't
learn it here either .. so I'd just remove it.
>
>
>>>>__read_mostly?
>>>
>>>No other xt_ modules have it, but I can add it.
>>
>>They don't? I'll add it to the others.
>
>
> Not even ipt_ or ip6t_ modules have it.
I've added it to all matches and tarets.
>
>
>>>The routing code (where Eric pointed me to) does it on boot-up.
>>
>>It does not use get_random_bytes during boot and reinitizalizes the
>>random value after a flush.
>
>
> So what should xt_connlimit do? Grab the rnd per xt_connlimit_data
> (= per rule)?
That doesn't seem necessary, just do what others do:
static int hash_rnd_initted;
static int hash_rnd;
static int hash(...)
{
if (unlikely(!hash_rnd_initted)) {
get_random_bytes(&hash_rnd, 4);
hash_rnd_initted = 1;
}
next prev parent reply other threads:[~2007-06-25 14:46 UTC|newest]
Thread overview: 71+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-06-20 9:39 xt_connlimit 20070620_2 Jan Engelhardt
2007-06-20 17:21 ` Andrew Beverley
2007-06-22 11:14 ` Patrick McHardy
2007-06-22 11:36 ` Jan Engelhardt
2007-06-22 11:43 ` Patrick McHardy
2007-06-22 12:33 ` Jan Engelhardt
2007-06-22 12:42 ` Patrick McHardy
2007-06-22 13:22 ` Jan Engelhardt
2007-06-22 13:27 ` Patrick McHardy
2007-06-25 11:11 ` xt_connlimit 20070625 Jan Engelhardt
2007-06-25 11:41 ` Patrick McHardy
2007-06-25 11:45 ` Patrick McHardy
2007-06-25 12:36 ` Jan Engelhardt
2007-06-25 12:43 ` Patrick McHardy
2007-06-25 14:41 ` Jan Engelhardt
2007-06-25 14:46 ` Patrick McHardy [this message]
2007-06-25 15:01 ` Jan Engelhardt
2007-06-25 15:05 ` Patrick McHardy
2007-06-25 15:05 ` Patrick McHardy
2007-06-25 15:14 ` Jan Engelhardt
2007-06-28 19:23 ` Jan Engelhardt
2007-06-28 19:27 ` Patrick McHardy
2007-06-28 19:31 ` Jan Engelhardt
2007-06-28 19:33 ` Patrick McHardy
2007-06-28 19:48 ` Patrick McHardy
2007-06-28 19:51 ` xt_connlimit 20070628 Jan Engelhardt
2007-06-28 19:55 ` xt_connlimit 20070628 kernel Jan Engelhardt
2007-06-29 11:27 ` Patrick McHardy
2007-07-01 14:11 ` Jan Engelhardt
2007-07-02 12:27 ` Patrick McHardy
2007-07-02 15:38 ` Jan Engelhardt
2007-07-02 15:40 ` Patrick McHardy
2007-07-02 19:53 ` Jan Engelhardt
2007-07-03 11:14 ` Patrick McHardy
2007-07-03 11:31 ` Jan Engelhardt
2007-07-03 11:34 ` Patrick McHardy
2007-07-04 10:56 ` Jan Engelhardt
2007-07-04 14:52 ` Patrick McHardy
2007-07-04 15:11 ` Jan Engelhardt
2007-07-06 13:05 ` Patrick McHardy
2007-07-07 17:51 ` xt_connlimit 20070707 kernel Jan Engelhardt
2007-07-09 14:30 ` Patrick McHardy
2007-07-09 15:10 ` Jan Engelhardt
2007-07-09 15:20 ` Patrick McHardy
2007-07-09 15:29 ` Patrick McHardy
2007-07-09 15:32 ` Jan Engelhardt
2007-07-09 15:33 ` Patrick McHardy
2007-07-09 15:34 ` Patrick McHardy
2007-07-09 15:38 ` Jan Engelhardt
2007-07-09 15:43 ` Patrick McHardy
2007-07-13 14:18 ` Yasuyuki KOZAKAI
[not found] ` <200707131418.l6DEIudN010879@toshiba.co.jp>
2007-07-13 15:01 ` Jan Engelhardt
2007-07-13 15:03 ` Patrick McHardy
2007-07-13 15:13 ` Jan Engelhardt
2007-07-13 15:16 ` Patrick McHardy
2007-07-13 15:31 ` Jan Engelhardt
2007-07-13 15:42 ` Patrick McHardy
2007-07-13 16:08 ` Jan Engelhardt
2007-07-13 15:44 ` Yasuyuki KOZAKAI
[not found] ` <200707131544.l6DFivSf011446@toshiba.co.jp>
2007-07-13 16:09 ` Jan Engelhardt
2007-07-10 6:30 ` Yasuyuki KOZAKAI
2007-07-11 17:37 ` Jan Engelhardt
2007-07-11 18:04 ` Patrick McHardy
2007-07-11 18:18 ` Jan Engelhardt
2007-07-11 18:19 ` Jan Engelhardt
2007-07-11 18:25 ` Patrick McHardy
[not found] ` <200707100630.l6A6UBM1021597@toshiba.co.jp>
2007-07-11 13:23 ` Patrick McHardy
2007-07-04 8:55 ` xt_connlimit 20070628 kernel Yasuyuki KOZAKAI
2007-07-04 14:52 ` Patrick McHardy
2007-06-28 20:08 ` xt_connlimit 20070628 Jan Engelhardt
2007-06-25 18:51 ` xt_connlimit 20070620_2 Patrick McHardy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=467FD52E.4050705@trash.net \
--to=kaber@trash.net \
--cc=jengelh@computergmbh.de \
--cc=netfilter-devel@lists.netfilter.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.