On Fri, 4 Aug 2023 at 18:26, Peter Maydell <peter.maydell@xxxxxxxxxx> wrote: > > On Thu, 27 Jul 2023 at 08:31, Akihiko Odaki <akihiko.odaki@xxxxxxxxxx> wrote: > > > > kvm_arch_get_default_type() returns the default KVM type. This hook is > > particularly useful to derive a KVM type that is valid for "none" > > machine model, which is used by libvirt to probe the availability of > > KVM. > > > > For MIPS, the existing mips_kvm_type() is reused. This function ensures > > the availability of VZ which is mandatory to use KVM on the current > > QEMU. > > > > Signed-off-by: Akihiko Odaki <akihiko.odaki@xxxxxxxxxx> > > --- > > include/sysemu/kvm.h | 2 ++ > > target/mips/kvm_mips.h | 9 --------- > > accel/kvm/kvm-all.c | 4 +++- > > hw/mips/loongson3_virt.c | 2 -- > > target/arm/kvm.c | 5 +++++ > > target/i386/kvm/kvm.c | 5 +++++ > > target/mips/kvm.c | 2 +- > > target/ppc/kvm.c | 5 +++++ > > target/riscv/kvm.c | 5 +++++ > > target/s390x/kvm/kvm.c | 5 +++++ > > 10 files changed, 31 insertions(+), 13 deletions(-) > > > > diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h > > index 115f0cca79..ccaf55caf7 100644 > > --- a/include/sysemu/kvm.h > > +++ b/include/sysemu/kvm.h > > @@ -369,6 +369,8 @@ int kvm_arch_get_registers(CPUState *cpu); > > > > int kvm_arch_put_registers(CPUState *cpu, int level); > > > > +int kvm_arch_get_default_type(MachineState *ms); > > + > > New global functions should have a doc comment that explains > what they do, what their API is, etc. For instance, is > this allowed to return an error, and if so, how ? Looks like this was the only issue with this patchset. So I propose to take this into my target-arm queue for 8.2, with the following doc comment added: /** * kvm_arch_get_default_type: Return default KVM type * @ms: MachineState of the VM being created * * Return the default type argument to use in the * KVM_CREATE_VM ioctl when creating the VM. This will * only be used when the machine model did not specify a * type to use via the MachineClass::kvm_type method. * * Returns: type to use, or a negative value on error. */ unless anybody wants more time for review or for this series to go into the tree some other way. thanks -- PMM
