On 07.06.2018 13:42, Paolo Bonzini wrote: > On 07/06/2018 13:27, Thomas Huth wrote: >>> As to "-enable-kvm", I don't see anything wrong with users using it, or >>> even with occasionally adding more options like it. However, we should >>> warn developers that such simple options should be syntactic sugar for a >>> structured (i.e. QemuOpts-based) option like "-accel", and that it >>> should only be done for similarity with existing options. >> Honestly, in this case I think it's just confusing for the normal users, >> and not sugar (anymore). If I'm an unexperienced user who wants to >> enable KVM, and I see multiple options that seem to be related, I wonder >> whether they do the same or whether there's a difference, and which one >> is preferred. And "-accel kvm" is even less to type than "-enable-kvm", >> so there is really no advantage for "-enable-kvm" anymore. I think we >> should remove "-enable-kvm" and "-enable-hax" from qemu-doc.texi and >> only list it in the new legacy chapter / document. > > Well, there's also the issue of distros shipping qemu-kvm binaries. I > think those should be provided by upstream. If we do that, then we're > perhaps in a better position to place --enable-kvm under the rug.
You don't need -enable-kvm for those qemu-kvm binaries, only -no-kvm. And -no-kvm has never been documented in our qemu-doc user documentation anyway (apart from the fact that it is now mentioned in the deprecation chapter). So if we'd now mentioned -no-kvm in a "legacy feature" chapter instead of the deprecation chapter, that would even improve the situation for downstream qemu-kvms :-) Thomas