| To: | Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: bad TSO performance in 2.6.9-rc2-BK |
| From: | "David S. Miller" <davem@xxxxxxxxxxxxx> |
| Date: | Thu, 23 Sep 2004 16:41:49 -0700 |
| Cc: | ak@xxxxxxx, niv@xxxxxxxxxx, andy.grover@xxxxxxxxx, anton@xxxxxxxxx, netdev@xxxxxxxxxxx |
| In-reply-to: | <E1CAdCy-0001FM-00@gondolin.me.apana.org.au> |
| References: | <20040923161141.4ea9be4c.davem@davemloft.net> <E1CAdCy-0001FM-00@gondolin.me.apana.org.au> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
On Fri, 24 Sep 2004 09:41:04 +1000 Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> wrote: > David S. Miller <davem@xxxxxxxxxxxxx> wrote: > > > > That means as the window fills up, we have to see 44 ACKs > > before we are able to send the next TSO frame. Needless to > > say that breaks ACK clocking completely. > > Hang on a second, the same problem would occur before the congestion > changes were made, right? Previously we counted TSO frames as single "packets" as long as we could fit one more frame into the congestion window we'd spit out the whole TSO frame, and this kept the pipe full. |
| Previous by Date: | Re: 2.6.9-rc2-mm2 fn_hash_insert oops, Martin Diehl |
|---|---|
| Next by Date: | Re: bad TSO performance in 2.6.9-rc2-BK, Herbert Xu |
| Previous by Thread: | Re: bad TSO performance in 2.6.9-rc2-BK, Herbert Xu |
| Next by Thread: | Re: bad TSO performance in 2.6.9-rc2-BK, Herbert Xu |
| Indexes: | [Date] [Thread] [Top] [All Lists] |