On Thu, Oct 30, 2014 at 01:49:22PM +0100, Markus Armbruster wrote: > Kevin Wolf <kw...@redhat.com> writes: > > > Am 29.10.2014 um 14:54 hat Markus Armbruster geschrieben: > >> Anthony tried something similar (commit 79368c8), but couldn't get it > >> right (commit 8b33d9e). > > > > The discussion back then: http://patchwork.ozlabs.org/patch/58980/ > > > > The problem with Anthony's code was that he didn't handle a qiov > > correctly that had unaligned members. With today's block layer, this is > > not a big deal to implement correctly. We're running coroutines instead > > of AIO callbacks and we don't have to do all the manual qiov fixing > > magic that Anthony had in his patch, util/iov.c provides all you need. > > > > I'll send out an RFC series that implements this. > > I'm strongly opposed to this idea.
Me too. Meddling with the guest is wrong and just adds a new problem! It is perfectly okay for backup appliances, nested virtualization, or disk management appliances to write image format headers to raw disks. I'm am 100% against trying to detect the guest writing image format headers to raw disks. Probing is not a guest problem. It is a QEMU problem. Fix QEMU, don't cripple the guest. Stefan
pgpbO0t7U9yOz.pgp
Description: PGP signature
