Re: [PATCH 1/4] KVM: support VT-d device hotplug

[Avi Kivity <avi@redhat.com>]

Han, Weidong wrote:
> From bba614bf2acf22f765995fb2364de04cec039226 Mon Sep 17 00:00:00 2001
> From: Weidong Han <weidong.han@intel.com>
> Date: Fri, 14 Nov 2008 16:53:10 +0800
> Subject: [PATCH] support VT-d device hotplug
>
> wrap kvm_assign_device() and kvm_deassign_device() to support assign/deassign a device to a guest
>   

> --- a/include/linux/kvm.h
> +++ b/include/linux/kvm.h
> @@ -425,6 +425,8 @@ struct kvm_trace_rec {
>  				   struct kvm_assigned_pci_dev)
>  #define KVM_ASSIGN_IRQ _IOR(KVMIO, 0x70, \
>  			    struct kvm_assigned_irq)
> +#define KVM_DEASSIGN_PCI_DEVICE _IOR(KVMIO, 0x71, \
> +				     struct kvm_assigned_pci_dev)
>  
>   

Need a KVM_CAP_ to indicate this is available.

>  
>  static inline void kvm_guest_enter(void)
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 4f43abe..689d615 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -293,7 +293,11 @@ static int kvm_vm_ioctl_assign_device(struct kvm *kvm,
>  	list_add(&match->list, &kvm->arch.assigned_dev_head);
>  
>  	if (assigned_dev->flags & KVM_DEV_ASSIGN_ENABLE_IOMMU) {
> -		r = kvm_iommu_map_guest(kvm, match);
> +		if (kvm->arch.intel_iommu_domain)
> +			r = kvm_assign_device(kvm, match);
> +		else
> +			r = kvm_iommu_map_guest(kvm, match);
> +
>   

Shouldn't kvm_assign_device() be called unconditionally?  Perhaps some 
code needs to be removed from kvm_iommu_map_guest().


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