Re: [PATCH] iommu/vt-d: avoid unnecessory panic if iommu init fail in tboot

[Zhenzhong Duan]

Hi Baolu,

On Mon, Nov 9, 2020 at 11:15 AM Lu Baolu <baolu.lu@xxxxxxxxxxxxxxx> wrote:
>
> Hi Zhenzhong,
>
> On 11/9/20 10:27 AM, Zhenzhong Duan wrote:
> > +intel iommu maintainers,
> >
> > Can anyone help review this patch? Thanks
> >
> > Zhenzhong
> >
> > On Wed, Nov 4, 2020 at 4:15 PM Zhenzhong Duan <zhenzhong.duan@xxxxxxxxx> wrote:
> >>
> >> "intel_iommu=off" command line is used to disable iommu and iommu is force
> >> enabled in a tboot system. Meanwhile "intel_iommu=tboot_noforce,off"
> >> could be used to force disable iommu in tboot for better performance.
> >>
> >> By default kernel should panic if iommu init fail in tboot for security
> >> reason, but it's unnecessory if we use "intel_iommu=tboot_noforce,off".
> >>
> >> Fix it by return 0 in tboot_force_iommu() so that force_on is not set.
> >>
> >> Fixes: 7304e8f28bb2 ("iommu/vt-d: Correctly disable Intel IOMMU force on")
> >> Signed-off-by: Zhenzhong Duan <zhenzhong.duan@xxxxxxxxx>
> >> ---
> >>   arch/x86/kernel/tboot.c | 5 +----
> >>   1 file changed, 1 insertion(+), 4 deletions(-)
> >>
> >> diff --git a/arch/x86/kernel/tboot.c b/arch/x86/kernel/tboot.c
> >> index 992fb1415c0f..9fd4d162b331 100644
> >> --- a/arch/x86/kernel/tboot.c
> >> +++ b/arch/x86/kernel/tboot.c
> >> @@ -511,12 +511,9 @@ struct acpi_table_header *tboot_get_dmar_table(struct acpi_table_header *dmar_tb
> >>
> >>   int tboot_force_iommu(void)
> >>   {
> >> -       if (!tboot_enabled())
> >> +       if (!tboot_enabled() || intel_iommu_tboot_noforce)
> >>                  return 0;
> >>
> >> -       if (intel_iommu_tboot_noforce)
> >> -               return 1;
>
> This was obviously wrong. It should return false, right?

I guess you missed above change: "if (!tboot_enabled() ||
intel_iommu_tboot_noforce)".
It does return false.

>
> It looks odd that intel_iommu_tboot_noforce is defined in Intel iommu
> implementation, but is used here. How about moving it back to the iommu
> driver?

That's better, will do. Thanks for your suggestion.

Zhenzhong