Re: [PATCH net-next] xsk: remove doubled defines

[Maciej Fijalkowski <maciej.fijalkowski@intel.com>]

On Thu, Mar 12, 2026 at 01:39:47PM +0100, Paolo Abeni wrote:
> On 3/10/26 10:34 PM, Stanislav Fomichev wrote:
> > On 03/10, Maciej Fijalkowski wrote:
> >> Seems we have been carrying around doubled defines for unaligned mode
> >> logic. Remove one of them.
> >>
> >> Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> >> ---
> >>  include/net/xsk_buff_pool.h | 7 -------
> >>  1 file changed, 7 deletions(-)
> >>
> >> diff --git a/include/net/xsk_buff_pool.h b/include/net/xsk_buff_pool.h
> >> index 0b1abdb99c9e..ccb3b350001f 100644
> >> --- a/include/net/xsk_buff_pool.h
> >> +++ b/include/net/xsk_buff_pool.h
> >> @@ -174,13 +174,6 @@ static inline void xp_dma_sync_for_device(struct xsk_buff_pool *pool,
> >>  	dma_sync_single_for_device(pool->dev, dma, size, DMA_BIDIRECTIONAL);
> >>  }
> >>  
> >> -/* Masks for xdp_umem_page flags.
> >> - * The low 12-bits of the addr will be 0 since this is the page address, so we
> >> - * can use them for flags.
> >> - */
> >> -#define XSK_NEXT_PG_CONTIG_SHIFT 0
> >> -#define XSK_NEXT_PG_CONTIG_MASK BIT_ULL(XSK_NEXT_PG_CONTIG_SHIFT)
> >> -
> >>  static inline bool xp_desc_crosses_non_contig_pg(struct xsk_buff_pool *pool,
> >>  						 u64 addr, u32 len)
> >>  {
> >> -- 
> >> 2.43.0
> >>
> > 
> > All consumers seem to be in net/xdp/xsk_buff_pool.c, 
> 
> AFAICS there are users in xsk_buff_pool.h, too:
> 
> $ git grep XSK_NEXT_PG_CONTIG_MASK include/
> include/net/xsk_buff_pool.h:#define XSK_NEXT_PG_CONTIG_MASK
> BIT_ULL(XSK_NEXT_PG_CONTIG_SHIFT)
> include/net/xsk_buff_pool.h:    xskb->frame_dma = (dma_pages[addr >>
> PAGE_SHIFT] & ~XSK_NEXT_PG_CONTIG_MASK) +
> include/net/xsk_buff_pool.h:#define XSK_NEXT_PG_CONTIG_MASK
> BIT_ULL(XSK_NEXT_PG_CONTIG_SHIFT)
> include/net/xsk_buff_pool.h:           !(pool->dma_pages[addr >>
> PAGE_SHIFT] & XSK_NEXT_PG_CONTIG_MASK);
> 
> > so why not remove the defines from xsk.h instead?
> 
> Indeed it looks like the macro definition is duplicated multiple times.
> @Maciej: could you please drop all the unneeded dups?

Thanks, let me send a v2 with a better cleanup:)

> 
> Thanks,
> 
> Paolo
> 
>