[Intel-wired-lan] [PATCH v4 net-next 8/9] ixgbe: don't reserve excessive XDP_PACKET_HEADROOM on XSK Rx to skb

Penigalapati, Sandeep sandeep.penigalapati at intel.com
Tue Jan 11 11:51:46 UTC 2022


>-----Original Message-----
>From: Intel-wired-lan <intel-wired-lan-bounces at osuosl.org> On Behalf Of
>Alexander Lobakin
>Sent: Wednesday, December 8, 2021 7:37 PM
>To: intel-wired-lan at lists.osuosl.org
>Cc: Song Liu <songliubraving at fb.com>; Jesper Dangaard Brouer
><hawk at kernel.org>; Daniel Borkmann <daniel at iogearbox.net>; Yonghong
>Song <yhs at fb.com>; Martin KaFai Lau <kafai at fb.com>; John Fastabend
><john.fastabend at gmail.com>; Alexei Starovoitov <ast at kernel.org>; Andrii
>Nakryiko <andrii at kernel.org>; Björn Töpel <bjorn at kernel.org>;
>netdev at vger.kernel.org; Jakub Kicinski <kuba at kernel.org>; KP Singh
><kpsingh at kernel.org>; bpf at vger.kernel.org; David S. Miller
><davem at davemloft.net>; linux-kernel at vger.kernel.org
>Subject: [Intel-wired-lan] [PATCH v4 net-next 8/9] ixgbe: don't reserve
>excessive XDP_PACKET_HEADROOM on XSK Rx to skb
>
>{__,}napi_alloc_skb() allocates and reserves additional NET_SKB_PAD
>+ NET_IP_ALIGN for any skb.
>OTOH, ixgbe_construct_skb_zc() currently allocates and reserves additional
>`xdp->data - xdp->data_hard_start`, which is XDP_PACKET_HEADROOM for
>XSK frames.
>There's no need for that at all as the frame is post-XDP and will go only to the
>networking stack core.
>Pass the size of the actual data only to __napi_alloc_skb() and don't reserve
>anything. This will give enough headroom for stack processing.
>
>Fixes: d0bcacd0a130 ("ixgbe: add AF_XDP zero-copy Rx support")
>Signed-off-by: Alexander Lobakin <alexandr.lobakin at intel.com>
>Reviewed-by: Michal Swiatkowski <michal.swiatkowski at linux.intel.com>
>---
> drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
Tested-by: Sandeep Penigalapati <sandeep.penigalapati at intel.com>


More information about the Intel-wired-lan mailing list