[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH 2/4] intel_iommu: Support IR-only mode without DMA translatio
From: |
David Woodhouse |
Subject: |
Re: [PATCH 2/4] intel_iommu: Support IR-only mode without DMA translation |
Date: |
Mon, 14 Mar 2022 15:45:47 +0000 |
User-agent: |
Evolution 3.36.5-0ubuntu1 |
On Mon, 2022-03-14 at 11:24 -0400, Michael S. Tsirkin wrote:
> On Mon, Mar 14, 2022 at 02:25:42PM +0000, David Woodhouse wrote:
> > From: David Woodhouse <dwmw@amazon.co.uk>
> >
> > By setting none of the SAGAW bits we can indicate to a guest that DMA
> > translation isn't supported. Tested by booting Windows 10, as well as
> > Linux guests with the fix at https://git.kernel.org/torvalds/c/c40aaaac10
> >
> >
> > Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
> > Reviewed-by: Peter Xu <peterx@redhat.com>
> > Acked-by: Jason Wang <jasowang@redhat.com>
>
> this is borderline like a feature, but ...
It's the opposite of a feature — it's turning the feature *off* ;)
> >
> > @@ -3627,12 +3628,17 @@ static void vtd_init(IntelIOMMUState *s)
> > s->next_frcd_reg = 0;
> > s->cap = VTD_CAP_FRO | VTD_CAP_NFR | VTD_CAP_ND |
> > VTD_CAP_MAMV | VTD_CAP_PSI | VTD_CAP_SLLPS |
> > - VTD_CAP_SAGAW_39bit | VTD_CAP_MGAW(s->aw_bits);
> > + VTD_CAP_MGAW(s->aw_bits);
> > if (s->dma_drain) {
> > s->cap |= VTD_CAP_DRAIN;
> > }
> > - if (s->aw_bits == VTD_HOST_AW_48BIT) {
> > - s->cap |= VTD_CAP_SAGAW_48bit;
> > + if (s->dma_translation) {
> > + if (s->aw_bits >= VTD_HOST_AW_39BIT) {
> > + s->cap |= VTD_CAP_SAGAW_39bit;
> > + }
> > + if (s->aw_bits >= VTD_HOST_AW_48BIT) {
> > + s->cap |= VTD_CAP_SAGAW_48bit;
> > + }
> > }
> > s->ecap = VTD_ECAP_QI | VTD_ECAP_IRO;
> >
>
>
> ... this looks like you are actually fixing aw_bits < VTD_HOST_AW_39BIT,
> right? So maybe this patch is ok like this since it also fixes a
> bug. Pls add this to commit log though.
Nah, aw_bits cannot be < VTD_HOST_AW_39BIT. We explicitly check in
vtd_decide_config(), and only 39 or 48 bits are supported,
corresponding to 3-level or 4-level page tables.
The only time we'd want to *not* advertise 39-bit support is if we
aren't advertising DMA translation at all. Which is the (anti-)feature
introduced by this patch.
smime.p7s
Description: S/MIME cryptographic signature
[PATCH 3/4] intel_iommu: Only allow interrupt remapping to be enabled if it's supported, David Woodhouse, 2022/03/14
Re: [PATCH 1/4] target/i386: Fix sanity check on max APIC ID / X2APIC enablement, Igor Mammedov, 2022/03/16