Re: [PATCH 1/2] xen/virtio: Fix n_pages calculation in xen_grant_dma_map(unmap)_page()

[Oleksandr Tyshchenko <olekstysh@xxxxxxxxx>]

On Thu, Oct 6, 2022 at 1:24 PM Juergen Gross <jgross@xxxxxxxx> wrote:

Hello Juergen

[sorry for the possible format issues]

On 06.10.22 12:14, Oleksandr Tyshchenko wrote:
>
>
> On Thu, Oct 6, 2022 at 11:05 AM Juergen Gross <jgross@xxxxxxxx
> <mailto:jgross@xxxxxxxx>> wrote:
>
>     On 06.10.22 09:35, Xenia Ragiadakou wrote:
>
>
>
> Hello Xenia, Juergen
>
> [sorry for the possible format issues]
>
>      >
>      > On 10/5/22 20:48, Oleksandr Tyshchenko wrote:
>      >> From: Oleksandr Tyshchenko <oleksandr_tyshchenko@xxxxxxxx
>     <mailto:oleksandr_tyshchenko@xxxxxxxx>>
>      >>
>      >> Take page offset into the account when calculating the number of pages
>      >> to be granted.
>      >>
>      >> Signed-off-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@xxxxxxxx
>     <mailto:oleksandr_tyshchenko@xxxxxxxx>>
>      >> Fixes: d6aca3504c7d ("xen/grant-dma-ops: Add option to restrict memory
>     access
>      >> under Xen")
>      >> ---
>      >>   drivers/xen/grant-dma-ops.c | 5 +++--
>      >>   1 file changed, 3 insertions(+), 2 deletions(-)
>      >>
>      >> diff --git a/drivers/xen/grant-dma-ops.c b/drivers/xen/grant-dma-ops.c
>      >> index 8973fc1e9ccc..1998d0e8ce82 100644
>      >> --- a/drivers/xen/grant-dma-ops.c
>      >> +++ b/drivers/xen/grant-dma-ops.c
>      >> @@ -153,7 +153,7 @@ static dma_addr_t xen_grant_dma_map_page(struct device
>      >> *dev, struct page *page,
>      >>                        unsigned long attrs)
>      >>   {
>      >>       struct xen_grant_dma_data *data;
>      >> -    unsigned int i, n_pages = PFN_UP(size);
>      >> +    unsigned int i, n_pages = PFN_UP(offset + size);
>      >
>      > Here, why do we use PFN_UP and not XEN_PFN_UP?
>      > Also, since the variable 'n_pages' seems to refer to the number of grants
>      > (unless I got it all entirely wrong ...), wouldn't it be more suitable to
>     call
>      > explicitly gnttab_count_grant()?
>
>     Good point.
>
>
> +1
>
>
>     I think this will need another patch for switching grant-dma-ops.c to
>     use XEN_PAGE_SIZE and XEN_PAGE_SHIFT.
>
>
> +1
>
> I can create a separate patch for converting on top of this series, it would be
> nice to clarify one point.
>
> So I will convert PAGE_SIZE/PAGE_SHIFT to XEN_PAGE_SIZE/XEN_PAGE_SHIFT
> respectively (where appropriate).

Yes, that would be the idea.

> Should the PFN_UP be converted to XEN_PFN_UP *or* use
> gnttab_count_grant() explicitly? Personally I would prefer the former, but would
> also be ok with the latter.

I agree XEN_PFN_UP would be better, especially as XEN_PAGE_SIZE/XEN_PAGE_SHIFT
will be used in the same functions.

 

Thanks for the clarification.

 

Juergen

--

Regards,

Oleksandr Tyshchenko