| To: | netdev@xxxxxxxxxxx |
|---|---|
| Subject: | Re: Linux Hardware/Software IP Stack Integration for a TOE |
| From: | Francois Romieu <romieu@xxxxxxxxxx> |
| Date: | Fri, 30 Mar 2001 10:10:54 +0200 |
| In-reply-to: | <D55EFF49CC829E468BE958686EDE9FDE10921A@nc8220exchange.ral.lucent.com>; from gparrott@lucent.com on Thu, Mar 29, 2001 at 03:02:02PM -0500 |
| References: | <D55EFF49CC829E468BE958686EDE9FDE10921A@nc8220exchange.ral.lucent.com> |
| Sender: | owner-netdev@xxxxxxxxxxx |
| User-agent: | Mutt/1.2.5i |
Gregory Parrott <gparrott@xxxxxxxxxx> écrit : [...] > 3) Where is the best place to hook in? I could intercept sys_* calls or > I could hook in at the specific protocol (tcp, udp, raw). My feeling is > it will be a combination of the two. IMVVVHO hooking in inet_create (net/ipv4/af_inet.c): you register your own proto fields and you may decide to fallback to the normal one (if your connect operation doesn't succeed for example). That way the existing stack shouldn't need any modification. -- Ueimor |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | HELP: Why are redirected packets dropped?, Bjorn Hammarberg |
|---|---|
| Next by Date: | Re: bug in ipv6 extension header parsing??, Andi Kleen |
| Previous by Thread: | Linux Hardware/Software IP Stack Integration for a TOE, Gregory Parrott |
| Next by Thread: | help, Dhiman Barman |
| Indexes: | [Date] [Thread] [Top] [All Lists] |