Re: [PATCH] KVM x86: Handle hypercalls for assigned PCI devices

[Avi Kivity <avi@qumranet.com>]

Amit Shah wrote:
>
>>> +	if (is_error_page(host_page)) {
>>> +		printk(KERN_INFO "%s: gfn %p not valid\n",
>>> +		       __func__, (void *)page_gfn);
>>> +		r = -1;
>>>       
>> r = -1 is not really informative. Better use some meaningful error.
>>     
>
> The error's going to the guest. The guest, as we know, has already done a 
> successful DMA allocation. Something went wrong in the hypercall, and we 
> don't know why (bad page). Any kind of error here isn't going to be 
> intelligible to the guest anyway. It's mostly a host thing if we ever hit 
> this.
>
>   

If the guest is not able to handle it, why bother returning an error?  
Better to kill it.

But in any case, -1 is not a good error number.

>>> +	if (find_pci_pt_dev(&vcpu->kvm->arch.pci_pt_dev_head,
>>> +			    &pci_pt_info, 0, KVM_PT_SOURCE_ASSIGN))
>>> +		r++; /* We have assigned the device */
>>> +
>>> +	kunmap(host_page);
>>>       
>> better use atomic mappings here.
>>     
>
> We can't use atomic mappings for guest pages. They can be swapped out.
>   

kmap()ed pages can't be swapped out either.  The atomic in kmap_atomic() 
only refers to the context in which the pages are used.  Atmoic kmaps 
are greatly preferable to the nonatomic ones.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.