From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Fri, 17 Aug 2012 13:26:06 +0200 Subject: [Buildroot] [PATCH 11/36] package/qemu: new package In-Reply-To: <201208141903.45500.yann.morin.1998@free.fr> References: <1344815664-28138-1-git-send-email-yann.morin.1998@free.fr> <1344815664-28138-11-git-send-email-yann.morin.1998@free.fr> <20120814153403.1673236e@skate> <201208141903.45500.yann.morin.1998@free.fr> Message-ID: <20120817132606.1212d7f8@skate> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Le Tue, 14 Aug 2012 19:03:45 +0200, "Yann E. MORIN" a ?crit : > > Just curious, what's the use case for Qemu on the target? I would have > > seen Qemu as something useful on the host, but on the target? > > I want to build a libvirt-based system to host VMs, and I do not want > to use a complete distro for that : I do not care about the handy-dandy > init stuff, I don't care about networking auto-managing stuff, no fscking > desktop stuff... I just want a basic system that just runs a libvirt daemon, > and manages the VMs. > > I want to build a completly-static QEMU, and that's not possible on a > desktop/server distro that uses glibc, because static linking still pulls > the NSS libs. So, rather than creating a new 'framework', I just decided > to reuse something. ;-) > > I want to play! ;-) Ok, interesting. The only thing I'm a bit worried about is that this brings Buildroot outside of its "environment to cross-compile Linux systems for embedded platforms". Although it initially probably fits the "minimal Linux system for VM hosting" use case, I'm worried that down the road the needs may diverge (for example people using Buildroot for VM hosting environments will want package management, etc.). But maybe I simply shouldn't anticipate too much on this and just see how things move. However, on this patch set, I'd like to see at least a Ack from Peter on the general idea of merging Qemu for the target and all the required dependencies. If Peter is fine with the general idea, we can together review the implementation details, but the general idea needs to be Acked, me thinks. Of course, other people in the community are invited to raise their voice here as well. Thanks! Thomas -- Thomas Petazzoni, Free Electrons Kernel, drivers, real-time and embedded Linux development, consulting, training and support. http://free-electrons.com