net/octeontx2: allow vector processing packets not by 4
authorVamsi Attunuru <vattunuru@marvell.com>
Fri, 20 Dec 2019 13:02:52 +0000 (18:32 +0530)
committerFerruh Yigit <ferruh.yigit@intel.com>
Fri, 17 Jan 2020 18:46:26 +0000 (19:46 +0100)
commit14766d69202b819b63ee2fccc1a9206b0172bb5e
treed452cf5a6fe0830d53ecc67e8fe07151e67f6c09
parent9a6dde4cd5515bef2263948d5ef15aae4fd8a94d
net/octeontx2: allow vector processing packets not by 4

Current vector mode implementation floor-aligns pkt count
with NIX_DESCS_PER_LOOP and process that many packets.

Patch addresses the case where pkt count modulo NIX_DESCS_PER_LOOP
is non-zero, after the vector mode processing, scalar routine is
used to process if there are any leftover packets. Scalar routine
is also used when descriptor head is about to wrap and turn out to
be unaligned.

Signed-off-by: Vamsi Attunuru <vattunuru@marvell.com>
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
drivers/net/octeontx2/otx2_rx.c
drivers/net/octeontx2/otx2_tx.c