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

[Xenia Ragiadakou <burzalodowa@xxxxxxxxx>]

On 10/5/22 20:48, Oleksandr Tyshchenko wrote:
> From: Oleksandr Tyshchenko <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>
> 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()?
If the above questions have been already answered in the past, please 
ignore.
>         grant_ref_t grant;
>         dma_addr_t dma_handle;
>   
> @@ -185,7 +185,8 @@ static void xen_grant_dma_unmap_page(struct device *dev, dma_addr_t dma_handle,
>                                      unsigned long attrs)
>   {
>         struct xen_grant_dma_data *data;
> -       unsigned int i, n_pages = PFN_UP(size);
> +       unsigned long offset = dma_handle & (PAGE_SIZE - 1);
> +       unsigned int i, n_pages = PFN_UP(offset + size);
>         grant_ref_t grant;
>   
>   	if (WARN_ON(dir == DMA_NONE))
--
Xenia