Yes. Yoshifuji is right.The timeStamp should be
-- hundredths of seconds since an epoch
TimeTicks ::=
[APPLICATION 3]
IMPLICIT INTEGER (0..4294967295)
I will modify the patch, and resubmit it against Linux-2.6.0-test6-bk8
again.
Thanks
Shirley Ma
IBM Linux Technology Center
15300 SW Koll Parkway
Beaverton, OR 97006-6063
Phone: (503) 578-7638
FAX: (503) 578-3228
YOSHIFUJI Hideaki / 吉藤英明 <yoshfuji@xxxxxxxxxxxxxx>@oss.sgi.com on
10/07/2003 04:35:18 AM
Sent by: netdev-bounce@xxxxxxxxxxx
To: davem@xxxxxxxxxx
cc: kuznet@xxxxxxxxxxxxx, mashirle@xxxxxxxxxxxxxxxxxxxxxxx,
netdev@xxxxxxxxxxx, yoshfuji@xxxxxxxxxxxxxx
Subject: Re: [PATCH] Implementation for IPv6 MIB:ipv6AddressTable
In article <20031007041246.13aa0391.davem@xxxxxxxxxx> (at Tue, 7 Oct 2003
04:12:46 -0700), "David S. Miller" <davem@xxxxxxxxxx> says:
> On Tue, 7 Oct 2003 15:16:20 +0400 (MSD)
> kuznet@xxxxxxxxxxxxx wrote:
>
> > Actually, this does not depend on our implementation. If people
implement
> > a MIB, required precision of these times is prescribed in the rfc.
> > And this is definitely not jiffies.
>
> Right, Shirley what is it?
The type is TimeTicks (RFC2578) via TimeStamp (RFC2579).
Thus, uint32_t counter in 1/100 sec. unit.
--yoshfuji
|