| To: | Andi Kleen <ak@xxxxxx> |
|---|---|
| Subject: | Re: [PATCH] Avoid gettimeofday when not needed |
| From: | Ben Greear <greearb@xxxxxxxxxxxxxxx> |
| Date: | Wed, 31 Mar 2004 09:30:02 -0800 |
| Cc: | Ville Nuorvala <vnuorval@xxxxxxxxxx>, netdev@xxxxxxxxxxx |
| In-reply-to: | <20040331151047.GA6033@xxxxxxxxxxxxx> |
| Organization: | Candela Technologies |
| References: | <200403311201.i2VC13D19579@xxxxxxxxxxx> <Pine.LNX.4.58.0403311552070.19615@xxxxxxxxxxxxxxx> <20040331135449.GB62402@xxxxxxxxxxxxx> <Pine.LNX.4.58.0403311705460.19615@xxxxxxxxxxxxxxx> <20040331151047.GA6033@xxxxxxxxxxxxx> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
| User-agent: | Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/20040113 |
Andi Kleen wrote: On Wed, Mar 31, 2004 at 05:24:47PM +0300, Ville Nuorvala wrote:There just doesn't seem to be any general flag field in sk_buff that we might use at this moment. Should one be added, or should we just hide the flag inside the cb field using the likes of inet_skb_parm and inet6_skb_parm?There are no free bits indeed. I think ->cb is the best choice then.
Why not add a 32-bit field to the skb to handle various flag needs
going forward?
If space is very important, could use an uint16 for flags and consolidate
the ip_summed and cloned fields, which seem to be boolean fields.
unsigned char local_df,
cloned,
pkt_type,
ip_summed;
Ben
--
Ben Greear <greearb@xxxxxxxxxxxxxxx>
Candela Technologies Inc http://www.candelatech.com
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: route cache DoS testing and softirqs, Dipankar Sarma |
|---|---|
| Next by Date: | Re: [PATCH]dump interface IPv6 multicast/anycast addresses through netlink, Shirley Ma |
| Previous by Thread: | Re: [PATCH] Avoid gettimeofday when not needed, Andi Kleen |
| Next by Thread: | Re: [PATCH] Avoid gettimeofday when not needed, Andi Kleen |
| Indexes: | [Date] [Thread] [Top] [All Lists] |