On Mon, Feb 25, 2013 at 02:03:02AM +0100, Igor Mammedov wrote: > - since hyperv_* helper functions are used only in target-i386/kvm.c > move them there as static helpers > > Signed-off-by: Igor Mammedov <[email protected]> > Requestd-By: Eduardo Habkost <[email protected]>
I wonder if we could do this later, to make review and testing easier. We could simply keep the hv_* handling inside parse_featurestr() by now. I don't think this would block any of the CPU hotplug or CPU model probing/compatibility/reliability work we're doing. I mean: * I don't expect hv-* to appear on a machine-type compat_props array in the near feature. * I don't expect people to need to set per-CPU hv-* properties on device_add for CPU hotplug. So we could keep them as special cases on parse_featurestr(), and convert them to per-CPU properties only after we have the subclasses and CPU hotplug working. Considering that -cpu options will be translated to global properties, it will be trivial to keep compatibility with existing behavior of "-cpu hv_*=..." once we change them from static variables to per-CPU fields. > --- > target-i386/Makefile.objs | 2 +- > target-i386/cpu.c | 16 +++++++--- > target-i386/cpu.h | 7 +++++ > target-i386/hyperv.c | 64 > --------------------------------------------- > target-i386/hyperv.h | 45 ------------------------------- > target-i386/kvm.c | 36 ++++++++++++++++++------- > 6 files changed, 45 insertions(+), 125 deletions(-) > delete mode 100644 target-i386/hyperv.c > delete mode 100644 target-i386/hyperv.h > [...] -- Eduardo
