| To: | prashant s <sree_prash@xxxxxxxxx> |
|---|---|
| Subject: | Regarding NAPI performance |
| From: | Robert Olsson <Robert.Olsson@xxxxxxxxxxx> |
| Date: | Tue, 27 Apr 2004 09:54:28 +0200 |
| Cc: | netdev@xxxxxxxxxxx |
| In-reply-to: | <20040426184039.91977.qmail@web60210.mail.yahoo.com> |
| References: | <20040426184039.91977.qmail@web60210.mail.yahoo.com> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
prashant s writes:
> driver without NAPI (rtl8139too in 2.4.20 linux)
> showed 72Mbps and the NAPI driver (rtl8139too in
> 2.6.3) showed 94Mbps,both having their ring size at
> 8k.The packet size i used was same as before.
> why is this difference when the ring size is reduced.
Well a good question. I would expect non-NAPI to perform
equal to NAPI at low network load even w. small RX-rings.
Any difference interrupt delay? NAPI can live without
delay. Or some other driver difference? I have no experience
with rtl8139too myself.
Cheers.
--ro
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: variation in thruput w/ change in mtu on gige, Abhijit Karmarkar |
|---|---|
| Next by Date: | Re: tun device - bug or feature? WAS(Re: IMQ / new Dummy device post., jamal |
| Previous by Thread: | Regarding NAPI performance, prashant s |
| Next by Thread: | Update FarSync WAN driver in 2.4.25, Kevin Curtis |
| Indexes: | [Date] [Thread] [Top] [All Lists] |