On Thu, May 29, 2025 at 12:54:31PM -0700, Mina Almasry wrote:
> On Wed, May 28, 2025 at 8:11 PM Byungchul Park <byungchul@xxxxxx> wrote:
> >
> > To simplify struct page, the effort to separate its own descriptor from
> > struct page is required and the work for page pool is on going.
> >
> > To achieve that, all the code should avoid directly accessing page pool
> > members of struct page.
> >
> > Access ->pp_magic through struct netmem_desc instead of directly
> > accessing it through struct page in page_pool_page_is_pp(). Plus, move
> > page_pool_page_is_pp() from mm.h to netmem.h to use struct netmem_desc
> > without header dependency issue.
> >
> > Signed-off-by: Byungchul Park <byungchul@xxxxxx>
> > ---
> > include/linux/mm.h | 12 ------------
> > include/net/netmem.h | 14 ++++++++++++++
> > mm/page_alloc.c | 1 +
> > 3 files changed, 15 insertions(+), 12 deletions(-)
> >
> > diff --git a/include/linux/mm.h b/include/linux/mm.h
> > index 8dc012e84033..de10ad386592 100644
> > --- a/include/linux/mm.h
> > +++ b/include/linux/mm.h
> > @@ -4311,16 +4311,4 @@ int arch_lock_shadow_stack_status(struct task_struct *t, unsigned long status);
> > */
> > #define PP_MAGIC_MASK ~(PP_DMA_INDEX_MASK | 0x3UL)
> >
> > -#ifdef CONFIG_PAGE_POOL
> > -static inline bool page_pool_page_is_pp(struct page *page)
> > -{
> > - return (page->pp_magic & PP_MAGIC_MASK) == PP_SIGNATURE;
> > -}
> > -#else
> > -static inline bool page_pool_page_is_pp(struct page *page)
> > -{
> > - return false;
> > -}
> > -#endif
> > -
> > #endif /* _LINUX_MM_H */
> > diff --git a/include/net/netmem.h b/include/net/netmem.h
> > index f05a8b008d00..9e4ed3530788 100644
> > --- a/include/net/netmem.h
> > +++ b/include/net/netmem.h
> > @@ -53,6 +53,20 @@ NETMEM_DESC_ASSERT_OFFSET(pp_ref_count, pp_ref_count);
> > */
> > static_assert(sizeof(struct netmem_desc) <= offsetof(struct page, _refcount));
> >
> > +#ifdef CONFIG_PAGE_POOL
> > +static inline bool page_pool_page_is_pp(struct page *page)
> > +{
> > + struct netmem_desc *desc = (__force struct netmem_desc *)page;
> > +
>
> Is it expected that page can be cast to netmem_desc freely? I know it
> works now since netmem_desc and page have the same layout, but how is
> it going to continue to work when page is shrunk and no longer has
This should be updated once struct netmem_desc has its own instance from
slab. As Pavel mentioned, that should be done another way.
> 'pp_magic' inside of it? Is that series going to fixup all the places
> where casts are done?
>
> Is it also allowed that we can static cast netmem_desc to page?
>
> Consider creating netmem_desc_page helper like ptdesc_page.
Do we need casting netmem_desc to page?
>
> I'm not sure the __force is needed too.
Ah, I will remove it.
Byungchul
>
> --
> Thanks,
> Mina
