From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Hall Subject: Re: Appropriate DPDK data structures for TCP sockets Date: Thu, 12 Mar 2015 23:41:03 -0700 Message-ID: <20150313064103.GA9389@mhcomputing.net> References: <3ABAA9DB-3F71-44D4-9C46-22933F9F30F0@mhcomputing.net> <20150222160204.20816910@urahara> <20150223211645.GB20766@mhcomputing.net> <54EBA0F2.6040409@cloudius-systems.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "" To: Avi Kivity Return-path: Content-Disposition: inline In-Reply-To: <54EBA0F2.6040409-RmZWMc9puTNJc61us3aD9laTQe2KTcn/@public.gmane.org> List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces-VfR2kkLFssw@public.gmane.org Sender: "dev" On Mon, Feb 23, 2015 at 11:51:46PM +0200, Avi Kivity wrote: > https://github.com/cloudius-systems/seastar Hi Avi and others, My code unintentionally ended up looking somewhat like a C version of your seastar C++ code, even though I didn't really look at yours too much when I coded mine as it was using a lot of hardcore C++ features I really don't have any clue about. :) Someday maybe we can all do a bake-off of tests of DPDK TCPs and Host TCPs and see what kind of stability, features, and performance we can get. I didn't use anything too crazy or high performance just yet... just rte_hash and some spinlocks... but I'm keeping all the collective advice in mind to figure out how I'm going to make all this stuff work right soon. Matthew.