[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [v3][PATCH 04/16] xen/passthrough: extend hypercall to support rdm reservation policy



> From: Chen, Tiejun
> Sent: Friday, June 12, 2015 2:31 PM
> 
> >> @@ -1899,7 +1900,8 @@ static int intel_iommu_add_device(u8 devfn, struct 
> >> pci_dev
> >> *pdev)
> >>                PCI_BUS(bdf) == pdev->bus &&
> >>                PCI_DEVFN2(bdf) == devfn )
> >>           {
> >> -            ret = rmrr_identity_mapping(pdev->domain, 1, rmrr);
> >> +            ret = rmrr_identity_mapping(pdev->domain, 1, rmrr,
> >> +                                        XEN_DOMCTL_DEV_RDM_RELAXED);
> >
> > Why did you hardcode relax policy here? Shouldn't the policy come
> > from hypercall flag?
> 
> I just saw we have one path to use intel_iommu_add_device(),
> 
> pci_add_device()
>      |
>      + if ( !pdev->domain )
>        {
>          pdev->domain = hardware_domain;
>          ret = iommu_add_device(pdev);
>           |
>           + hd->platform_ops->add_device()
>               |
>               + intel_iommu_add_device()
> 
> So I think intel_iommu_add_device() is used to add a device to
> hardware_domain. And in our case hardware_domain should be special as I
> explained above.

Then please add a clear comment in such case.

Thanks
Kevin

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.