Re: [PATCH v2 4/6] mm: introduce page_offline_(begin|end|freeze|thaw) to synchronize setting PageOffline()

[Mike Rapoport]

On Fri, May 14, 2021 at 07:22:45PM +0200, David Hildenbrand wrote:
> A driver might set a page logically offline -- PageOffline() -- and
> turn the page inaccessible in the hypervisor; after that, access to page
> content can be fatal. One example is virtio-mem; while unplugged memory
> -- marked as PageOffline() can currently be read in the hypervisor, this
> will no longer be the case in the future; for example, when having
> a virtio-mem device backed by huge pages in the hypervisor.
> 
> Some special PFN walkers -- i.e., /proc/kcore -- read content of random
> pages after checking PageOffline(); however, these PFN walkers can race
> with drivers that set PageOffline().
> 
> Let's introduce page_offline_(begin|end|freeze|thaw) for
> synchronizing.
> 
> page_offline_freeze()/page_offline_thaw() allows for a subsystem to
> synchronize with such drivers, achieving that a page cannot be set
> PageOffline() while frozen.
> 
> page_offline_begin()/page_offline_end() is used by drivers that care about
> such races when setting a page PageOffline().
> 
> For simplicity, use a rwsem for now; neither drivers nor users are
> performance sensitive.
> 
> Acked-by: Michal Hocko <mhocko@xxxxxxxx>
> Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>

One nit below, otherwise

Reviewed-by: Mike Rapoport <rppt@xxxxxxxxxxxxx>

> ---
>  include/linux/page-flags.h | 10 ++++++++++
>  mm/util.c                  | 40 ++++++++++++++++++++++++++++++++++++++
>  2 files changed, 50 insertions(+)
> 
> diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
> index daed82744f4b..ea2df9a247b3 100644
> --- a/include/linux/page-flags.h
> +++ b/include/linux/page-flags.h
> @@ -769,9 +769,19 @@ PAGE_TYPE_OPS(Buddy, buddy)
>   * relies on this feature is aware that re-onlining the memory block will
>   * require to re-set the pages PageOffline() and not giving them to the
>   * buddy via online_page_callback_t.
> + *
> + * There are drivers that mark a page PageOffline() and do not expect any

Maybe "and expect there won't be any further access"...

> + * further access to page content. PFN walkers that read content of random
> + * pages should check PageOffline() and synchronize with such drivers using
> + * page_offline_freeze()/page_offline_thaw().
>   */
>  PAGE_TYPE_OPS(Offline, offline)
>  
> +extern void page_offline_freeze(void);
> +extern void page_offline_thaw(void);
> +extern void page_offline_begin(void);
> +extern void page_offline_end(void);
> +
>  /*
>   * Marks pages in use as page tables.
>   */
> diff --git a/mm/util.c b/mm/util.c
> index a8bf17f18a81..a034525e7ba2 100644
> --- a/mm/util.c
> +++ b/mm/util.c
> @@ -1010,3 +1010,43 @@ void mem_dump_obj(void *object)
>  }
>  EXPORT_SYMBOL_GPL(mem_dump_obj);
>  #endif
> +
> +/*
> + * A driver might set a page logically offline -- PageOffline() -- and
> + * turn the page inaccessible in the hypervisor; after that, access to page
> + * content can be fatal.
> + *
> + * Some special PFN walkers -- i.e., /proc/kcore -- read content of random
> + * pages after checking PageOffline(); however, these PFN walkers can race
> + * with drivers that set PageOffline().
> + *
> + * page_offline_freeze()/page_offline_thaw() allows for a subsystem to
> + * synchronize with such drivers, achieving that a page cannot be set
> + * PageOffline() while frozen.
> + *
> + * page_offline_begin()/page_offline_end() is used by drivers that care about
> + * such races when setting a page PageOffline().
> + */
> +static DECLARE_RWSEM(page_offline_rwsem);
> +
> +void page_offline_freeze(void)
> +{
> +	down_read(&page_offline_rwsem);
> +}
> +
> +void page_offline_thaw(void)
> +{
> +	up_read(&page_offline_rwsem);
> +}
> +
> +void page_offline_begin(void)
> +{
> +	down_write(&page_offline_rwsem);
> +}
> +EXPORT_SYMBOL(page_offline_begin);
> +
> +void page_offline_end(void)
> +{
> +	up_write(&page_offline_rwsem);
> +}
> +EXPORT_SYMBOL(page_offline_end);
> -- 
> 2.31.1
> 

-- 
Sincerely yours,
Mike.