Re: [PATCH 2/2] KVM: Fix kvm_free_physmem_slot memory leak.

[Avi Kivity <avi@redhat.com>]

(late reply, sorry)

François Diakhate wrote:
> [Sorry, I realized I forgot to check style, here is the fixed patch]
>
> Make sure that kvm_free_physmem_slot also frees the VM memory
> if it was allocated by the kernel.
>
>  /*
>   * Free any memory in @free but not in @dont.
>   */
> -static void kvm_free_physmem_slot(struct kvm_memory_slot *free,
> +static void kvm_free_physmem_slot(struct kvm *kvm,
> +				  struct kvm_memory_slot *free,
>  				  struct kvm_memory_slot *dont)
>  {
> +	if (!dont || free->userspace_addr != dont->userspace_addr) {
> +		struct kvm_userspace_memory_region mem = {
> +			.slot = memslot_id(kvm, free),
> +			.guest_phys_addr = free->base_gfn << PAGE_SHIFT,
> +			.memory_size = 0,
> +			.flags = 0,
> +		};
> +		kvm_arch_set_memory_region(kvm, &mem, *free, free->user_alloc);
> +	}
> +
>  	if (!dont || free->rmap != dont->rmap)
>   

What happens here if the both free and dont have nonzero, differnt 
->userspace_addr values?  Is is even possible?

Also, the call chain is fishy.  set_memory_region calls 
free_physmem_slot which calls arch_set_memory_region.  This is turning 
into pasta.


-- 
error compiling committee.c: too many arguments to function