Arnd
I apologize for the late reply. Many thanks for your patches.
1. I have committed your first patch for fixing timestamps on
unpatched kernels
2. concerning the patch below
- netif_rx() is called when NAPI is not used
- netif_receive_skb() is called when NAPI is used
So IMHO it cannot happen that your call twice ring_handler().
Can you please tell me what network adapter you're using and if its
drived is NAPI or not-NAPI enabled?
Thanks, Luca
On Jul 19, 2007, at 4:36 PM, Arnd Hannemann wrote:
Hi Luca,
Luca Deri schrieb:
Arnd
thanks for the report. Can you please investigate some of the issues
reported and provide patches?
Thanks Luca
Arnd Hannemann wrote:
2. Receiving packets are always captured twice
I investigated the 2nd issue and found out that ring_handler() is
called twice in the RX path:
in net/core/dev.c netif_rx()
and
in net/core/dev.c: netif_receive_skb()
I deleted the call in netif_receive_skb() and now capturing works
fine.
However I don't feel I understand the kernel code well enough to
submit a patch for this problem.
Best regards,
Arnd
_______________________________________________
Ntop-dev mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-dev
_______________________________________________
Ntop-dev mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-dev