| To: | jmorris@xxxxxxxxxxxxxxxx |
|---|---|
| Subject: | Re: [PATCH] LSM networking: skb hooks for 2.5.42 (2/7) |
| From: | "David S. Miller" <davem@xxxxxxxxxx> |
| Date: | Tue, 15 Oct 2002 10:40:14 -0700 (PDT) |
| Cc: | kuznet@xxxxxxxxxxxxx, netdev@xxxxxxxxxxx, linux-security-module@xxxxxxxxx |
| In-reply-to: | <Mutt.LNX.4.44.0210160027110.27676-100000@blackbird.intercode.com.au> |
| References: | <Mutt.LNX.4.44.0210160027110.27676-100000@blackbird.intercode.com.au> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
There is no way in hell that you are going to add a function call for every time we set the socket owner of a SKB. That is a critical path and it happens millions upon millions of times a second on a busy server. Performance will be penalized by changes like this. I want all this junk #ifdef'd if it is necessary, in particular the new member added to sk_buff et al. There is no way this stuff is going to happen in the default build, at least not in my networking tree :-) |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | [PATCH] LSM networking: tcp hooks for 2.5.42 (7/7), James Morris |
|---|---|
| Next by Date: | Re: [PATCH] LSM networking: tcp hooks for 2.5.42 (7/7), David S. Miller |
| Previous by Thread: | [PATCH] LSM networking: skb hooks for 2.5.42 (2/7), James Morris |
| Next by Thread: | Re: [PATCH] LSM networking: skb hooks for 2.5.42 (2/7), Donald Becker |
| Indexes: | [Date] [Thread] [Top] [All Lists] |