Ar Maw, 2006-08-22 am 20:29 +0200, ysgrifennodd Arjan van de Ven: > > And it doesn't work for VMI or lhype, both of which might modify > > paravirt_ops way later in the boot process, when loaded as a module. > > doesn't this then start to have the same issues that runtime patching > the system call table had? It has several I can see that are if anything worse - Stacked hypervisors stomping each others functions - Locking required to do updates: and remember our lock functions use methods in the array - If we boot patch inline code to get performance natively its almost impossible to then revert that.
