Re: [PATCH v2 11/18] AMD/IOMMU: return old PTE from {set,clear}_iommu_pte_present()

["Roger Pau Monné" <roger.pau@citrix.com>]

On Fri, Dec 10, 2021 at 01:59:02PM +0100, Jan Beulich wrote:
> On 10.12.2021 13:05, Roger Pau Monné wrote:
> > On Fri, Sep 24, 2021 at 11:51:40AM +0200, Jan Beulich wrote:
> >> In order to free intermediate page tables when replacing smaller
> >> mappings by a single larger one callers will need to know the full PTE.
> >> Flush indicators can be derived from this in the callers (and outside
> >> the locked regions). First split set_iommu_pte_present() from
> >> set_iommu_ptes_present(): Only the former needs to return the old PTE,
> >> while the latter (like also set_iommu_pde_present()) doesn't even need
> >> to return flush indicators. Then change return types/values and callers
> >> accordingly.
> > 
> > Without looking at further patches I would say you only care to know
> > whether the old PTE was present (ie: pr bit set), at which point those
> > functions could also return a boolean instead of a full PTE?
> 
> But looking at further patches will reveal that I then also need the
> next_level field from the old PTE (to tell superpages from page tables).

Oh, OK. I was expecting something like that.

Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>

I wouldn't mind if you added a note to the commit message that the
full PTE is returned because new callers will require more data.

Thanks, Roger.