Re: [PATCH 14/16] iommu/irq: Use amd_iommu_irq_ops if supported

[Joerg Roedel <joerg.roedel-5C7GfCeVMHo@public.gmane.org>]

On Fri, Sep 28, 2012 at 05:39:03PM -0600, Shuah Khan wrote:
> On Fri, Sep 28, 2012 at 6:24 AM, Joerg Roedel <joerg.roedel-5C7GfCeVMHo@public.gmane.org> wrote:
> >  void __init setup_irq_remapping_ops(void)
> >  {
> >         remap_ops = &intel_irq_remap_ops;
> > +
> > +#ifdef CONFIG_AMD_IOMMU
> > +       if (amd_iommu_irq_ops.prepare() == 0)
> > +               remap_ops = &amd_iommu_irq_ops;
> > +#endif
> 
> Should remap_ops be set to null when amd_iommu_irq_ops.prepare()
> fails?What happens if remap_ops left set to intel_irq_remap_ops?
> Should remap_ops = &intel_irq_remap_ops; moved into #else case
> forifdef CONFIG_AMD_IOMMU?

Remap-Ops does not need to be set to NULL because irq_remapping_enabled
will not get set to true then and remap_ops will not get called. The
Intel path can't also be moved to #else because this would mean that
kernels can only support eihter, Intel or AMD IOMMU. But Linux can
support both in the same kernel.


	Joerg

-- 
AMD Operating System Research Center

Advanced Micro Devices GmbH Einsteinring 24 85609 Dornach
General Managers: Alberto Bozzo
Registration: Dornach, Landkr. Muenchen; Registerger. Muenchen, HRB Nr. 43632