Date: Mon, 23 Aug 2004 20:40:51 +0200 From: Pawel Malachowski <pawmal-posting@freebsd.lublin.pl> To: Pawel Jakub Dawidek <pjd@freebsd.org> Cc: FreeBSD-gnats-submit@freebsd.org Subject: Re: bin/70311: ipfw(8) pipe/queue show N displays data not only for N Message-ID: <20040823184051.GC42452@shellma.zin.lublin.pl> In-Reply-To: <20040823165718.GQ30151@darkness.comp.waw.pl> References: <20040811163426.E06283474C2@shellma.zin.lublin.pl> <20040823165718.GQ30151@darkness.comp.waw.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Aug 23, 2004 at 06:57:18PM +0200, Pawel Jakub Dawidek wrote:
> This patch doesn't work for me as expected. It shows all pipes and queues
> with parentpipe=N.
> Could you check this patch (it is against -CURRENT, but it should be easy
> to apply it to 4.x):
Weird.
For me, Your patch gives on my working RELENG_4 shaper setup:
% ipfw pipe show 1
00001: unlimited 0 ms 12 KB 0 queues (1 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0x00000000/0x0000
q00110: weight 1 pipe 1 8 KB 0 queues (512 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0xffffffff/0x0000
q00120: weight 99 pipe 1 8 KB 0 queues (512 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0xffffffff/0x0000
// expected, show pipe 1 and every queue with parent=1
% ipfw queue show 1
00001: unlimited 0 ms 12 KB 0 queues (1 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0x00000000/0x0000
// not expected, there is no queue 1, but it shows pipe 1 here.
// if I want pipe 1, I would type ipfw pipe show 1
% ipfw pipe show 110
// expected, there is no pipe 110
% ipfw queue show 110
q00110: weight 1 pipe 1 8 KB 0 queues (512 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0xffffffff/0x0000
// expected, there is queue 110.
> Index: ipfw2.c
> ===================================================================
> RCS file: /private/FreeBSD/src/sbin/ipfw/ipfw2.c,v
> retrieving revision 1.54
> diff -u -p -r1.54 ipfw2.c
> --- ipfw2.c 12 Aug 2004 22:06:55 -0000 1.54
> +++ ipfw2.c 23 Aug 2004 16:51:26 -0000
> @@ -1564,6 +1564,12 @@ list_pipes(void *data, uint nbytes, int
> l = sizeof(*fs) + fs->rq_elements * sizeof(*q);
> next = (char *)fs + l;
> nbytes -= l;
> +
> + if (rulenum != 0 && ((rulenum != fs->fs_nr && do_pipe == 2) ||
> + (rulenum != fs->parent_nr && do_pipe == 1))) {
> + continue;
> + }
> +
> q = (struct dn_flow_queue *)(fs+1);
> sprintf(prefix, "q%05d: weight %d pipe %d ",
> fs->fs_nr, fs->weight, fs->parent_nr);
>
> Basically, it looks that first addition of 'do_pipe==2' is bogus.
Well, with this first 'do_pipe==2' check it works here like this:
% ipfw pipe show 1
00001: unlimited 0 ms 12 KB 0 queues (1 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0x00000000/0x0000
q00110: weight 1 pipe 1 8 KB 0 queues (512 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0xffffffff/0x0000
q00120: weight 99 pipe 1 8 KB 0 queues (512 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0xffffffff/0x0000
// expected, shows pipe 1 and every queue with parent=1
% ipfw queue show 1
// expected, there is no queue 1
% ipfw pipe show 110
// expected, there is no pipe 110
% ipfw queue show 110
q00110: weight 1 pipe 1 8 KB 0 queues (512 buckets) droptail
mask: 0x00 0x00000000/0x0000 -> 0xffffffff/0x0000
// expected, there is queue 110
So, for me it is not bogus. :)
cheers,
--
Paweł Małachowski
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040823184051.GC42452>