Jan Ceuleers wrote: > Arjan van de Ven wrote: >> Most offload features (TSO etc) end up helping (a little). Anything that >> makes the system more efficient will help; it's the "if you have less >> work to do >> you consume less power" concept. >> Exception tends to be TOE (total offloading), since TOE tends to be >> implemented >> by just another CPU, one on the card. This tends to be a loss in terms >> of power, >> because you now have 2 cpus slurping power, having just one tends to be >> more >> optimal. >> >> As for powertop suggesting this: I'm actually surprised these features >> arent >> on by default already. They should be.... >> If you have a specific one in mind, can you give some examples? It's easier >> to discuss such things on a specific case basis > > I have three machines with different ethernet chips in them. None of > them come with all of the offloading features that they support enabled > on boot. > > For example, the via-velocity in one machine supports only one offload > (which I now know is not hardware-assisted): > > [EMAIL PROTECTED]:~# ethtool -k eth0 > Offload parameters for eth0: > Cannot get device rx csum settings: Operation not supported > Cannot get device tx csum settings: Operation not supported > Cannot get device scatter-gather settings: Operation not supported > Cannot get device tcp segmentation offload settings: Operation not supported > Cannot get device udp large send offload settings: Operation not supported > rx-checksumming: off > tx-checksumming: off > scatter-gather: off > tcp segmentation offload: off > udp fragmentation offload: off > generic segmentation offload: off > > [EMAIL PROTECTED]:~# lspci -vv | grep Ethernet > 00:0e.0 Ethernet controller: VIA Technologies, Inc. VT6120/VT6121/VT6122 > Gigabit Ethernet Adapter (rev 11) > > Same with the natsemi chip in another machine: > > [EMAIL PROTECTED]:~# ethtool -k eth0 > Offload parameters for eth0: > Cannot get device rx csum settings: Operation not supported > Cannot get device tx csum settings: Operation not supported > Cannot get device scatter-gather settings: Operation not supported > Cannot get device tcp segmentation offload settings: Operation not supported > Cannot get device udp large send offload settings: Operation not supported > rx-checksumming: off > tx-checksumming: off > scatter-gather: off > tcp segmentation offload: off > udp fragmentation offload: off > generic segmentation offload: off > > [EMAIL PROTECTED]:~# lspci -vv | grep Ethernet > 00:06.0 Ethernet controller: National Semiconductor Corporation DP83815 > (MacPhyter) Ethernet Controller > > I cannot reboot the third machine (which has an r8169 device) at the > moment, but whereas it does support several hardware-assisted offload > features I seem to remember that only checksumming is enabled at boot > (but I can't remember whether it's rx or tx, but definitely not both). I > have since enabled all of the other features to produce the following: > > [EMAIL PROTECTED]:~# ethtool -k eth0 > Offload parameters for eth0: > Cannot get device udp large send offload settings: Operation not supported > rx-checksumming: on > tx-checksumming: on > scatter-gather: on > tcp segmentation offload: on > udp fragmentation offload: off > generic segmentation offload: on > > [EMAIL PROTECTED]:~# lspci -vv | grep Ethernet > 00:09.0 Ethernet controller: Realtek Semiconductor Co., Ltd. > RTL-8110SC/8169SC Gigabit Ethernet (rev 10) > > All of the above machines run Ubuntu 7.10 with a 2.6.22-based kernel.
Those features should IMHO be enabled by default in any case. I would almost suggest sending a patch upstream to enable these features by default. If there is a good reason to _not_ do this, then I'm sure you'll hear that from the maintainers. I personally do not know these drivers nor hardware well enough. Auke _______________________________________________ Power mailing list [email protected] http://www.bughost.org/mailman/listinfo/power
