[Top] [All Lists]

Re: [PATCH] pkt_cls.h incompatiables

To: hadi@xxxxxxxxxx
Subject: Re: [PATCH] pkt_cls.h incompatiables
From: YOSHIFUJI Hideaki / 吉藤英明 <yoshfuji@xxxxxxxxxxxxxx>
Date: Sat, 24 Jul 2004 21:33:23 -0400 (EDT)
Cc: davem@xxxxxxxxxx, shemminger@xxxxxxxx, arekm@xxxxxxxxxxxxx, netdev@xxxxxxxxxxx, yoshfuji@xxxxxxxxxxxxxx
In-reply-to: <1090675993.1161.11.camel@xxxxxxxxxxxxxxxx>
Organization: USAGI Project
References: <20040723135945.7f50b02c.davem@xxxxxxxxxx> <1090630154.1134.6.camel@xxxxxxxxxxxxxxxx> <1090675993.1161.11.camel@xxxxxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx

In article <1090675993.1161.11.camel@xxxxxxxxxxxxxxxx> (at 24 Jul 2004 09:33:52 
-0400), jamal <hadi@xxxxxxxxxx> says:

> Makes the performance counters a separate TLV -  a little bit of a pain,
> but acceptable given its optional.
> Compiles but hasnt been tested; user space code will have to change.

basically, looks fine.

1. is it okay to use unsigned long?
   I'd rather prefer __u32 or __u64.

2. please use kcnts[0] instead of kcnts[] because
   user probably want to do sizeof(struct tc_u32_pcnt)

> I dont want to touch the policer for now.

hmm. :-P


Hideaki YOSHIFUJI @ USAGI Project <yoshfuji@xxxxxxxxxxxxxx>
GPG FP: 9022 65EB 1ECF 3AD1 0BDF  80D8 4807 F894 E062 0EEA

<Prev in Thread] Current Thread [Next in Thread>