On 29/04/2015 18:08, Bandan Das wrote: >>>> >> > Yeah... I hear you. Ok, let me put it this way - Assume that we can >>>> >> > defer this allocation up until the point that the nested subsystem is >>>> >> > actually used i.e L1 tries running a guest and we try to allocate this >>>> >> > area. If get_free_page() failed in that case, would we still want to >>>> >> > kill L1 too ? I guess no. >>> >> >>> >> We could block the hypervisor thread on the allocation, just like it >>> >> would block on faults for swapped out pages or new ones that have to be >>> >> reclaimed from the page cache first. > So, block on a failure hoping that eventually it will succeed ? > >> > In that case we should avoid making the allocation GFP_ATOMIC to begin with. >> > >> > If a GFP_KERNEL allocation failed, returning -ENOMEM from KVM_RUN (which >> > practically means killing the guest) would actually be a very real >> > possibility. > Sorry Paolo, I missed your point. Isn't the allocation already GFP_KERNEL ? I mean if it were done lazily as in your thought-experiment. Then: - a GFP_ATOMIC allocation would be bad - a GFP_KERNEL allocation would block like Jan said; if it failed, I would be okay with returning -ENOMEM to userspace, even if that in practice means killing the guest. Paolo -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
