| To: | jamal <hadi@xxxxxxxxxx> |
|---|---|
| Subject: | Re: [PATCH]: Cruft fixes |
| From: | Andi Kleen <ak@xxxxxx> |
| Date: | Sat, 20 May 2000 10:23:52 +0200 |
| Cc: | davem@xxxxxxxxxx, kuznet@xxxxxxxxxxxxx, Andi Kleen <ak@xxxxxx>, SteveW@xxxxxxx, netdev@xxxxxxxxxxx |
| In-reply-to: | <Pine.GSO.4.20.0005192219010.17134-101000@shell.cyberus.ca>; from jamal on Sat, May 20, 2000 at 04:23:31AM +0200 |
| References: | <Pine.GSO.4.20.0005192219010.17134-101000@shell.cyberus.ca> |
| Sender: | owner-netdev@xxxxxxxxxxx |
On Sat, May 20, 2000 at 04:23:31AM +0200, jamal wrote: > > > This patch removes a bit of cruft. Please double check. Sorry, my original judgement was wrong. The RX path uses it to distingush between old protocols that don't support SMP multithreading yet and new multithreaded protocols. Unfortunately this breaks the fastroute packet socket detection hack as you noted. The right fix is to add a flag word to packet_type that distingushes packet sockets and non locked protocols, and use that instead of the overloaded data member. Such a flag word would fit nicely between type and dev, there should be a 2 bytes gap there anyways. -Andi |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH]: Cruft fixes, Jeff Garzik |
|---|---|
| Next by Date: | Re: tx_timeout and timer serialisation, Donald Becker |
| Previous by Thread: | Re: [PATCH]: Cruft fixes, Jeff Garzik |
| Next by Thread: | Re: [PATCH]: Cruft fixes, jamal |
| Indexes: | [Date] [Thread] [Top] [All Lists] |