> -----Original Message----- > From: Andi Kleen [mailto:[EMAIL PROTECTED] > Sent: Sunday, August 21, 2005 9:36 AM > To: David S. Miller > Cc: Leonid Grossman; [EMAIL PROTECTED]; [EMAIL PROTECTED]; > [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED]; > [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED] > Subject: Re: [PATCH] TCP Offload (TOE) - Chelsio
> The flow cache would also be very useful to allow the NIC to > send a MSI-X only to the CPU the user process did last recvmsg on. > This way the receiving could be all CPU local and packets > would need to be passed between CPUs. The process will likely > stay on the same CPU due to CPU affinity, so this should work > on average. > > This would need some state in the NIC, but it is much easier > than full TOE because it doesn't have to be exact and is just > a hint. If NIC and stack get desynchronized in some corner > cases all that happens is that receiving goes slightly > slower, but no data corruption can happen. > > I think that could be done with a single callback to the > netdevice from when TCP frees the skb. Andi, can you provide a callback patch please? As I mentioned in Ottawa, MSI-X and Receive Traffic Distribution patches to s2io driver is actually something that we are going to submit soon; this will also fit nicely in multi-queue support and NIC virtualization efforts that are accelerating in the industry as a whole. Multi-queue support is a separate big topic, this is the feature that I/O AT got absolutely right and it will be good to see if it could be utilized/supported by modules like NAPI and bonding. Anyways, the driver can make RTD decisions on it's own but I'd much rather rely on a clue from the host :-) Leonid > > -Andi > > - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
