From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [RFC/GIT PULL] Linux KVM tool for v3.2 Date: Fri, 4 Nov 2011 09:02:25 -0400 Message-ID: <20111104130225.GA24563@infradead.org> References: <20111104121601.GA15206@infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Christoph Hellwig , Linus Torvalds , Avi Kivity , Andrew Morton , Ingo Molnar , linux-kernel@vger.kernel.org, kvm@vger.kernel.org To: Pekka Enberg Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On Fri, Nov 04, 2011 at 02:35:18PM +0200, Pekka Enberg wrote: > We are reusing kernel code and headers and I am not interested in > copying them over. Living in the kernel tree is part of the design, > whether you like it or not. That's pretty much a blanko argument for throwing everything into the kernel. What's next, throwing your jvm into the kernel because you like some kernel helpers? We've been through this a few times - there is no reason why a tool using the KVM ioctls should be considered close to the kernel - all other users get away just fine staying out of tree, and that has helped to keep the ABI stable. This sounds more like you should create a libkernelutil with useful data structures we use in the kernel for userspace programs. I'd love to use that for a few projects, btw.