| To: | Ben Greear <greearb@xxxxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: [e1000 2.6 10/11] TxDescriptors -> 1024 default |
| From: | "David S. Miller" <davem@xxxxxxxxxx> |
| Date: | Thu, 11 Sep 2003 16:29:11 -0700 |
| Cc: | jgarzik@xxxxxxxxx, scott.feldman@xxxxxxxxx, netdev@xxxxxxxxxxx, ricardoz@xxxxxxxxxx |
| In-reply-to: | <3F6103BB.5030706@xxxxxxxxxxxxxxx> |
| References: | <Pine.LNX.4.44.0309081953510.1261-100000@xxxxxxxxxxxxxxxxxxxxx> <3F60CA6D.9090503@xxxxxxxxx> <3F60D0F3.8080006@xxxxxxxxxxxxxxx> <20030911131219.0ab8dfdd.davem@xxxxxxxxxx> <3F60DDCC.5020906@xxxxxxxxxxxxxxx> <20030911140746.4f0384a1.davem@xxxxxxxxxx> <3F60E947.4090005@xxxxxxxxxxxxxxx> <20030911142906.74d9dfe5.davem@xxxxxxxxxx> <3F60F3F7.6090203@xxxxxxxxxxxxxxx> <20030911160252.6cd6c07d.davem@xxxxxxxxxx> <3F6103BB.5030706@xxxxxxxxxxxxxxx> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
On Thu, 11 Sep 2003 16:22:35 -0700 Ben Greear <greearb@xxxxxxxxxxxxxxx> wrote: > David S. Miller wrote: > > So, again, dampen the per-socket send queue sizes. > > That's just a band-aid to cover up the flaw with the lack > of queue-pressure feedback to the higher stacks, as would be increasing the > TxDescriptors for that matter. The whole point of the various packet scheduler algorithms are foregone if we're just going to queue up and send the crap again. |
| Previous by Date: | Re: [e1000 2.6 10/11] TxDescriptors -> 1024 default, Ben Greear |
|---|---|
| Next by Date: | Re: [RESEND 1/3] /proc/net/{igmp,msfilter,raw,rt_cache,ip6_flowlabel,msfilter6,raw6} may drop some data, David S. Miller |
| Previous by Thread: | Re: [e1000 2.6 10/11] TxDescriptors -> 1024 default, Ben Greear |
| Next by Thread: | Re: [e1000 2.6 10/11] TxDescriptors -> 1024 default, jamal |
| Indexes: | [Date] [Thread] [Top] [All Lists] |