On 18/12/18 00:26, Jim Mattson wrote: > On Mon, Dec 3, 2018 at 1:54 PM Sean Christopherson > <sean.j.christopherson@xxxxxxxxx> wrote: > >> diff --git a/arch/x86/kvm/vmx/vmcs12.c b/arch/x86/kvm/vmx/vmcs12.c >> new file mode 100644 >> index 000000000000..53dfb401316d >> --- /dev/null >> +++ b/arch/x86/kvm/vmx/vmcs12.c > Won't all of these "new" files cause tremendous pain for backports to > the LTS branches? Shouldn't arch/x86/kvm/vmx/vmcs12.c, for instance, > look to git like a semantic move of arch/x86/kvm/vmx.c, followed by a > bunch of deletions, rather than a new file constructed from thin air? Yes, it will be. What you propose however won't work because git doesn't really track copy+delete in a different way than just creating a new file. It will just be up to the maintainers to provide backports for particularly important bugfixes. Paolo
