bff6a98f74
Avoids recalculation of DPAA2_IOVA_TO_VADDR, since this is somewhat costly function on no-IOMMU setups. Brings about 0.5 Gbps improvement at 19 Gbps speeds on LX2160 with 512b packets. Signed-off-by: Roman Kapl <rka@sysgo.com> Reviewed-by: Hemant Agrawal <hemant.agrawal@nxp.com>